CAMPBELL, Circuit Judge.
Forbro Design Corporation appeals from a judgment of the district court, 390 F. Supp. 794 (D. Mass. 1975), dismissing its patent infringement action against Raytheon Company on the ground that Forbro's patent, No. 3,028,538 (more particularly its claims 1-4) was obvious under 35 U.S.C. § 103. Raytheon cross-appeals from the court's dismissal of Raytheon's counterclaim alleging that Forbro had violated the antitrust laws in its use of the patent, and from the court's rejection of its affirmative defenses other than obviousness and of its claim for attorneys' fees.1
The patent in suit was applied for on August 4, 1958, by Aaron Rosenfeld and Kenneth Kupferberg and issued on April 3, 1962. Thereafter it was assigned to Forbro, of which Dr. Kupferberg is a principal. Though the original patent had seven claims, Raytheon is only charged with infringing claims 1-4.
The patent is for a transistorized power supply circuit providing direct current voltage that is automatically regulated to remain constant. Sophisticated electronic gear, such as computers, require exceedingly accurate and stable sources of power. To meet their needs, circuits like the patented circuit were developed that even out fluctuations in the voltage coming from an unregulated source.
Through Forbro's parent company, Kepco, Inc., Kenneth Kupferberg, the inventor, a physicist, and his three brothers, have manufactured regulated power supplies of their own design since the 1940's. Up to the time of the patented circuit, their power supplies employed vacuum tubes, but by late 1956, transistors were economically feasible, and Dr. Kupferberg became interested in developing transistorized power supplies. Although transistors were well known after the early 1950's and Kupferberg was not the first to apply them to power supplies, they had been too costly before the mid-50's for use in most commercial circuits.
On March 1, 1957, Kupferberg hired Robert Moog, a young engineer with some background in transistors, to assist him in developing a transistorized power supply. Moog was instructed that they had only four or five months to design and build a marketable model, and this schedule was met. Proceeding from Kepco's previously developed vacuum tube models, especially Model 2600, Moog in consultation with Kupferberg2 engineered the patented circuit, allegedly by June 7, 1957. Kupferberg is said to have sketched the basic concept in early March, though whether by March 16 as stated in certain interference papers or at another time is unclear.
The alleged invention is a "combination" invention - that is, the design of the circuit, not any novelty in its components, is what is claimed to be unique. In 1957 all the basic ingredients were well known. The circuit employs a control-amplifier, a device which compares variations in output voltage against a stable reference voltage, and reacts to any inequality by developing a control voltage that is used to restore equality. It also employs a series-pass device which, reacting to the correction from the control amplifier, varies its resistance in such a manner as to offset and balance the change in the dc output voltage. Dr. Kupferberg testified that from 1948 on Kepco manufactured power supplies capable of regulating all necessary voltages and employing series-pass elements and a reference.
The patented circuit was conceived, according to Dr. Kupferberg, with several objectives. One was the minimization of voltage across the control amplifier and the dissipation of energy in the series-pass device. Another was to achieve a "universal" circuit - one that could use the same control amplifier and the same components with different output voltages. Previous devices, due in part to the fact that higher voltages were required to operate the vacuum tubes, had had to be designed specifically for different voltage ranges. Dr. Kupferberg testified that the key to achieving his objectives was the patented way of interconnecting the transistorized components. A model utilizing the new circuit was displayed publicly in August, 1957. The basic concept has since been embodied in thousands of units produced commercially by Kepco, and has been employed by others, including Raytheon.
The utility of the circuit is not questioned. What is questioned, by Raytheon, is whether designing the circuit required more skill or knowledge than one might expect of qualified persons presented with the opportunity to use transistors and having Kupferberg's aims in mind.
Forbro claims that Kupferberg's insight led to "a new theory of design", and it points especially to the connection of one input terminal of the control amplifier, one terminal of the load circuit, one terminal of the reference voltage, and one terminal of the series-pass device at a common point, terminal 21 in the patent. This arrangement is said to permit a very small reference voltage and to keep the voltage across the control amplifier very low. Thus 1000 volts can be controlled by a reference of but one or two volts. Forbro contends that this basic design, featuring great flexibility and simplicity, has been used without change ever since it was developed, demonstrating its unique value.
The district court found that the only prior art cited in the patent was Chase, U.S. Patent No. 2,751,549, and that neither party raised this as pertinent at trial. Instead, attention at trial focused particularly upon U.S. Patent No. 2,840,777 ("DeBlasio")3 which, Raytheon claimed, fully anticipated the patent in suit. Raytheon also contends that appellant's own unpatented Model 2600, using tubes, largely anticipated the patented circuit.
The court found that circuitry revealed in DeBlasio was not identical, so as to render the patent in suit void under § 102. However, it concluded that the teaching of DeBlasio, added to what would have been known to one with ordinary skill in the art in 1957, would have enabled the designing of a "common point" and standardized control amplifier, two principal advantages claimed for the patented circuit. A third claimed advantage, the low voltage reference, was held not to have been a significant aim of the inventor. The court stated that while the patent in suit had enjoyed commercial success, there was no showing that it solved a problem that had remained unsolved.
"There was evidence that other power supply manufacturers were engaged in the transistorization of power supplies and the stipulations in Appendix 'B' reveal that the technology to arrive at the answers contained in the patent was available in the prior art." 390 F. Supp. at 802.
The court accordingly held the patent invalid for obviousness under 35 U.S.C. § 103.
35 U.S.C. § 103 provides that a patent may not be obtained, even though the invention is not disclosed or otherwise anticipated under § 102,
"if 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."
This codification replaced, in 1952, the previous "invention" standard. See Graham v. John Deere, 383 U.S. 1, 14-15, 15 L. Ed. 2d 545, 86 S. Ct. 684 (1966). Underlying both formulations is the constitutionally-based premise that no one may hold a patent monopoly if the effect is "to remove existent knowledge from the public domain, or to restrict free access to materials already available." Id. at 6; A. & P. Tea Co. v. Supermarket Corp., 340 U.S. 147, 152, 95 L. Ed. 162, 71 S. Ct. 127 (1950). Thus here the district court had to decide whether the patented circuitry represented an obvious application of knowledge already within the public domain, or whether the concept of the patentee added measurably to the public's store of knowledge, and so entitled him to the reward of a patent.
We turn now to Forbro's assignments of error.
1. Forbro insists that the district court was wrong to reject its contention that because DeBlasio and Model 2600 were both cited in the file of an interference proceeding taking place soon after the patent was allowed, the patent is entitled to an "enhancement" of the presumption of validity. Hildreth v. Mastoras, 257 U.S. 27, 32, 66 L. Ed. 112, 42 S. Ct. 20 (1921); Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 430 F.2d 221 (6th Cir. 1970), cert. denied, 401 U.S. 939, 28 L. Ed. 2d 218, 91 S. Ct. 932 (1971); Reynolds v. Whitin Machine Works, 167 F.2d 78 (4th Cir.) cert. denied, 334 U.S. 844, 92 L. Ed. 1768, 68 S. Ct. 1513 (1948); Gordon Form Lathe Co. v. Walcott Machine Co., 32 F.2d 55, 57 (6th Cir. 1929). The district court ruled that the interference was not concerned with the patentability of claims; that there was no enhancement; and that the fact that some pertinent prior art was not before the examiner in the initial prosecution "somewhat weakens the statutory presumption."4
The parties weave intricate arguments and counter-arguments that, on the one hand, patentability was a direct concern of the primary examiner who ruled on the motions to dissolve the interference and, on the other, that he was "powerless" to evaluate the patentability of Kupferberg's claims. Several facts are reasonably clear: the primary examiner did discover and study DeBlasio; he did rule that claim 5 of the subject patent was fully anticipated by DeBlasio; inexplicably he did not make a similar ruling as to claim 6, a transistorization of claim 5, suggesting that he did not give equal attention to the patentability of all claims. Under Rule 237 of the Patent Office, 37 C.F.R. § 1.237, the primary examiner may take note of unpatentability, but it is not clear either in theory or in practice how far he must go, or is likely to go, in reviewing patentability de novo. While we might infer that the primary examiner did not think claims 1-4 were anticipated by DeBlasio (the same view reached by the district court), we have little assurance as to the depth of his inquiry into the patentability of those claims under § 103, an issue presenting a wider range of considerations than whether one patent "reads" on another. Graham v. John Deere, supra. We conclude, as did the district court, that such Patent Office review as can be ascertained did not result in the kind of "enhancement" described in Hildreth v. Mastoras, supra, a case in which there were far greater indicia of Patent Office study of the specific litigated issues. Indeed, we do not believe the later, limited examination fully made up for the incompleteness of the prior art at the earlier application stage. We therefore find no error in the district court's conclusion of a "somewhat weakened" statutory presumption.
2. Forbro's next two objections relate to Raytheon's expert, Knutrud. Because the court spoke of his being qualified in "circuit design" rather than "the regulated power supply field," Forbro accuses it of overlooking the precept in Graham v. John Deere, supra, that obviousness must relate to the pertinent art. Alternatively, it argues that because Knutrud was not fully skilled in the power supply field in 1957, he could not give an opinion as to the level of ordinary skill in the art at the time. Both these arguments are, on this record, close to frivolous. It is perfectly clear that the court understood it was concerned with power supply circuits, not radio telescopes or sonic depthfinders. Appellant admits as much. And Knutrud, an M.I.T. graduate and Westinghouse employee, had considerable experience in voltage regulator circuits even by 1957, though with some modesty, as well as accuracy, he did not consider himself an "expert", apparently meaning authority, then. Expertise for legal purposes means that a witness has sufficient specialized knowledge, skill, experience, training, or education to testify in the form of an opinion. Fed. R. Evid. 702. The trial court has wide discretion in determining when a purported expert is sufficiently qualified to take the stand and render an opinion in a certain area. Salem v. United States Lines Co., 370 U.S. 31, 35, 8 L. Ed. 2d 313, 82 S. Ct. 1119 (1962). To be sure, there are limits to discretion; and a person lacking qualifications in the pertinent art or some other basis for testifying as to what would or would not have been known by one of ordinary skill in the pertinent art, should not be allowed to express an opinion. But by 1957 Knutrud had worked with numerous circuits using voltage regulators and transistors, and had designed a few, though not many, power supply units. By the time he testified, in 1974, his expertise was manifest and was indeed, unchallenged. In permitting him to testify as to the level of ordinary skill in 1957, the court could take account both of his present expertise and the working proficiency he had attained by 1957 in the relevant area. There was no abuse of discretion in receiving his testimony, the weight of which, of course, Forbro was entitled to contest and to seek to diminish through cross-examination.
3. Forbro's remaining arguments against the finding of § 103 invalidity are aimed at alleged factual misconceptions below. We thus turn briefly to the weight to be given the district court's findings. Arguably, while subsidiary findings stand unless clearly erroneous,5 the ultimate question - obviousness itself - is a "legal" question; and perhaps, in some cases, it is possible so to isolate the legal standard from its factual bedding. See, e.g., A. & P. Tea Co. v. Supermarket Corp., supra. More often, however, such as here, obviousness as an ultimate question cannot meaningfully be separated from those factual determinations which are peculiarly within the trial court's province, such as the credibility of the experts. The district court's supported findings on obviousness will therefore normally stand unless manifesting a misconception of the correct legal standard. See Shanklin Corp. v. Springfield Photo Mount Co., 521 F.2d 609, 616 (1st Cir. 1975), cert. denied, 424 U.S. 914, 96 S. Ct. 1112, 47 L. Ed. 2d 318, 44 U.S.L.W. 3451 (U.S. 1975); Koppers Company v. Foster Grant Co., 396 F.2d 370, 372 (1st Cir. 1968) (Aldrich, C.J.).
Appellant argues that the district court disregarded "the essential requirement that there be a suggestion in the prior art to make the inventive combination by one skilled in the pertinent art." But while an unobvious combination may later, "by infallible hindsight", seem more obvious than it was, the present record supports a finding that the combination itself, and not merely its constituent parts, would have been obvious to one skilled in the regulated power supply art. See Sinclair Co. v. Interchemical Corp., 325 U.S. 327, 89 L. Ed. 1644, 65 S. Ct. 1143 (1945); Dempster Brothers, Inc. v. Buffalo Metal Container Corp., 352 F.2d 420, 422-23 (2d Cir. 1965). Section 103 "does not require that each precise element of the claim be found in the prior art as appellant contends, but requires rather that we inquire as to what one of ordinary skill in this particular art would learn from the references." Application of Reynaud, 51 C.C.P.A. 1310, 331 F.2d 625, 627 (1964).
Knutrud testified that from col. 6, lines 14-30, of DeBlasio he could have readily devised the patented circuit, and he presented sketches showing how this could be done. Appellant points to Knutrud's concessions that DeBlasio was not explicit as to a common point, and to several other of his concessions on cross-examination. However, it was up to the district court to assess his entire testimony and weigh any inconsistencies; and Knutrud testified elsewhere that if he were building DeBlasio, replacing tubes with transistors, "there probably would be no difference from the patented circuit." Relocating the pass element was, he said, "a minor engineering change" and one with the skill of an artisan in 1957 would have accomplished the patented design. Given this and other testimony, including that of Kupferberg and Barber on cross-examination, and stipulations 21, 23, 27 and 28 (set forth at the conclusion of the district court's opinion), we could not, without invading the province of the trier of fact, reject the district court's finding of § 103 invalidity.
Moreover, the relevant secondary factors - what Judge Learned Hand called "sign posts"6 - tend to corroborate the decision below. There is no evidence that the patented circuit solved a long-standing problem that had defied solution. Rather the circuit was quickly formulated in response to the decision to transistorize previous tube circuits. Such a changeover required a degree of redesigning because transistors could not handle high voltages and had certain other properties not possessed by tubes, such as an ability to function at very low voltages. But it is not claimed that transistorization itself entailed patentable invention; and here, once he decided to use transistors, Dr. Kupferberg moved with assurance to the circuit in question. There was no extended period during which he or others are shown either to have failed to achieve a transistorized power supply, or else to have achieved one having defects which the present design remedied. Indeed, in affidavits filed in the early 60's in the interference (when it was to his advantage to establish priority of invention), Dr. Kupferberg maintained that in 1951 and 1953 he had already designed an early form of a "bridge" circuit and had placed the series pass in the negative lead. The patented circuit was presented in that document as a "transistorization" of Model 2600. Dr. Kupferberg thus opened himself to Raytheon's present insistence that the patented circuit was but a step along a well worn path.
None of the secondary factors set forth in the preceeding paragraph by themselves rule out the validity of the patent. They are consistent, however, with the conclusion that the patented circuit was an obvious application of already familiar technology. We cannot say that the district court exceeded proper limits in making that assessment.
4. We also find adequate record support for the court's statement that insertion of a PNP pass transistor in DeBlasio results in the "common point" advantage, that it would provide a circuit with the advantage of a standardized control amplifier, and that it would be obvious to one with ordinary skill in the art. This seems to us to be a reasonable if simplistic layman's interpretation of Knutrud's testimony, although Knutrud did not expressly refer to stipulation 28 as such. There is, to be sure, the testimony of others to the contrary, but we do not find it apparent that the court below erred in crediting Knutrud's views.
5. Finally, appellant argues that the court erred in giving no weight to the small voltage reference as a difference setting the patented device apart from the prior art. The court noted that Dr. Kupferberg had deposed that the inventive concept was contained in the first few paragraphs of the patent, and that nowhere was the benefit of a small voltage reference mentioned.
It may be, as appellant contends, that an unexpected property of an invention, if otherwise relevant, should not be ignored because unforeseen by the inventor. But we do not see how the fact that this circuit could be used with a small voltage reference helps appellant over the court's finding that the circuit itself was an obvious derivation from DeBlasio and existing technical knowledge. If the circuit was obvious to begin with, the fact that it had various desirable properties would not make it less so.
We conclude that the district court did not err in finding claims 1-4 to be invalid under § 103, and accordingly affirm its judgment for Raytheon on Forbro's claim of infringement.
There remains for consideration Raytheon's cross appeal from the court's adverse rulings on its affirmative defenses, its antitrust counterclaim, and its claim for counsel fees.
As we affirm the ruling of § 103 invalidity, we do not reach, and do not consider, the district court's disposition of the "on sale" defense under 35 U.S.C. § 102(b); the failure-to-file defense under 35 U.S.C. § 135(c); and the fraud defense. These matters were treated by the court, and by the cross-appellant in its appeal, solely as affirmative defenses to Forbro's complaint for infringement; and they are now immaterial in light of our affirmance of the dismissal under § 103 of Forbro's infringement claim.
Raytheon's antitrust claim was designated a counterclaim. It was ordered dismissed on the merits by the district court as the court found no "willful" or "intentional" fraud sufficient to establish a violation of section 2 of the Sherman Act, 15 U.S.C. section 2. On appeal, Raytheon asserts not only the section 2 claim but seeks to make out a claim under section 1 of the Sherman Act. However, the section 1 claim was not pleaded by Raytheon, the district court having denied Raytheon's motion, made on the eve of trial, to include the claim by amendment. Raytheon neither appealed from the denial, nor has pointed out any abuse of discretion in the court's ruling, and we decline to entertain the section 1 claim.
As for the section 2 claim, Raytheon introduced no proof of "the exclusionary power of the illegal patent claim in terms of the relevant market for the product involved", necessary to make out a section 2 violation. Walker Process Equipment, Inc. v. Food Machinery & Chemical Corp., 382 U.S. 172, 177, 15 L. Ed. 2d 247, 86 S. Ct. 347 (1965). The proof which Raytheon did produce on, e.g., market dominance, went to patent-related issues which involve different considerations than do antitrust claims, see id. at 177-78, and there was no showing that in this case the questions were the same. Thus even apart from the district court's findings as to lack of willful or intentional fraud, we are satisfied that Raytheon failed to show a violation of section 2 of the Sherman Act.
Raytheon finally seeks attorneys' fees under 35 U.S.C. § 285. No such award was made below, and this is plainly not a case where we feel impelled to order the district court, against its better judgment, to make such an award.
The judgment of the district court is affirmed.
* Of the District of Maryland sitting by designation.
1. The court ruled that the patent, if valid, was infringed, and that it had not been anticipated under § 102, did not violate "on sale" time limitations of § 102(b), and was not unenforceable under § 135(c). The court discussed but declined to rule on Raytheon's claim of invalidity based on alleged fraud in procurement and during an interference proceeding.
2. Aaron Rosenfeld had nothing to do with the circuit, and his name was withdrawn as co-inventor in 1963. Rosenfeld had worked with Kupferberg, but only beginning in June, 1957. Kupferberg explains that when, in August, 1958, the present patent was sought, it was mistakenly believed that Rosenfeld had been involved.
3. DeBlasio first came up as a significant prior art reference in an interference that took place in the early 60's after issuance of the subject patent. The examiner determined that DeBlasio read on claim 5 of the subject patent, leading Kupferberg to abandon that claim. Forbro claims that DeBlasio is of little relevance to claims 1-4 in that it does not reveal the "common point" connection nor permit a lower voltage reference and the flexibility of the patented circuit.
4. Forbro points out that three of the same primary examiners who considered DeBlasio's patent application also considered the subject application. From this, Forbro asks us to infer that DeBlasio's status as prior art was, in fact, carefully weighed by the Patent Office. We are unwilling to indulge in this inference, in part because it is grounded upon pure speculation and in part because the premise is refuted by the allowance of claim 5 even though claim 5, as was later discovered, was fully anticipated by DeBlasio.
5. Fed. R. Civ. P. 52(a) provides: "Findings of fact shall not be set aside unless clearly erroneous, and due regard shall be given to the opportunity of the trial court to judge of the credibility of the witnesses."
6. "There are indeed some sign posts: e.g. how long did the need exist; how many tried to find the way; how long did the surrounding and accessory arts disclose the means; how immediately was the invention recognized as an answer by those who used the new variant?" Reiner v. I. Leon Co., 285 F.2d 501, 504 (2d Cir. 1960), cert. denied, 366 U.S. 929, 6 L. Ed. 2d 388, 81 S. Ct. 1649 (1961), quoted in Dempster Brothers, Inc. v. Buffalo Metal Container Corp., 352 F.2d 420, 422-23 (2d Cir. 1965).