This is an action for patent infringement brought by Harry W. Becker, plaintiff-appellant, against Webcor, Inc., defendant-appellee. The district court found the patents in question to be invalid and not infringed and rendered judgment in favor of Webcor. This appeal followed.

Plaintiff Becker is the owner of the two patents in suit, No. 2,595,443 (No. -443) filed March 14, 1946 and No. 2,595,444 (No. -444) filed June 26, 1946. Both patents were issued to Becker on May 6, 1952.

The two patents relate to electronic audio amplifiers and are primarily concerned with the electrical circuitry in the power output stage. No. -443 discloses a single tube circuit, the one principally considered in this action, in which all claims are in issue. No. -444 discloses a double tube circuit which shows the use of the Becker circuit of No. -443 in a conventional push-pull relationship, in which only claim 3 is in issue.

Plaintiff Becker is an individual residing in Skokie, Illinois and is self-employed as a consultant and designer of various types of radio and television equipment. Defendant Webcor is an Illinois corporation with a place of business in Chicago, Illinois and since May 6, 1952 has made, used and sold a wide variety of amplifier circuits.

Electronic audio amplifiers are a part of a well-developed art. It has been recognized for many years in the art that such amplifiers introduce distortion into the signal being amplified. The result has been that the output signal differs with respect to the input signal. This is due to the fact that vacuum tubes are not perfect devices and introduce into the signal various order harmonics which distort the output signal. This distortion produces an unpleasant effect upon the listener.

Stated in another way, harmonic distortion is that undesirable distortion of the signal being amplified which arises from the inefficiencies of the amplifier and is manifest as spurious frequencies which are the multiples of the frequency being amplified. Such distortion is reduced by inverting the phase of the distortion and feeding it back into the signal amplification path to cancel itself. As a result some loss of amplification of the main signal occurs.

The common purpose of all the circuits brought to the trial court's attention is to improve reproduction fidelity by the reduction of harmonic distortions occurring in the power stage of a signal amplifier.

There are two identifiable prior-art ways of reducing distortion. One is by the "push-pull" method used only in double tube amplifiers, and the other is by "inverse feedback," commonly referred to as "negative feedback," used in all types of amplifiers, both double tube and single tube.

One of the earlier disclosures of negative feedback was in the Black United States Patent No. 2,102,671 issued in 1937. Black's negative feedback consisted of taking a portion of the distorted output of the amplifier and reintroducing it into an input stage of the amplifier in such relationship to the input signal that the portion of the signal fed back tended to cancel a portion of the distortion normally occurring in the amplifier, as well as a portion of the fundamental signal, with the result that overall distortion was reduced.

The prior art disclosed the true pushpull circuit utilizing two separate vacuum tubes. In the push-pull circuit even harmonic distortion components of the current produced by the two tubes tend to cancel, while the fundamental and odd harmonics tend to add. Because the second harmonic component is by far the most powerful of all the distortion components, its cancellation effects a substantial reduction in overall distortion and a substantial improvement in amplifier output, while the extremely weak higher harmonic components do not materially affect the signal.

Becker claims that with the circuit he designed, as described in the patents in suit, he was able to effect a reduction in the most important source of distortion, the second harmonic, in a manner similar to that obtained with the two-tube push-pull circuit of the prior art without the high signal loss resulting from the use of negative feed-back as in the Black invention. He characterized the operation of his amplifier as producing a "push-pull effect," referred to in the trial as the "Becker effect."

Becker contends his No. -443 discloses and claims the original Becker circuit employing a single output tube. He urges that his No. -444 embodies the basic Becker concept of his No. -443, but further discloses its application to a true push-pull circuit using two amplifier tubes. He asserts that his inventions satisfied a long felt need in the highly developed and competitive amplifier industry by providing higher fidelity at low cost.

Becker charges Webcor infringed all claims of his No. -443 by its amplifier ex-emplified by Webcor Model 100-621; claims 1, 3 and 5 of No. -443 by amplifiers exemplified by Webcor Models 100-601 (and other listed models); and claim 3 of No. -444 by amplifiers exemplified by Webcor Models 333 (and other listed models).

The prior art relied upon at the trial was:

    United States Patents:
      Black Patent No. 2,102,671, issued
      December 21, 1937
      Cunningham Patent No. 2,429,124,
      issued October 14, 1947
    Foreign Patents:
      Swiss Patent No. 216,071, issued
      November 1, 1941
      French Patent No. 806,775, issued
      October 5, 1936
      British Patent No. 541,260 issued
      November 15, 1941
    Publication:
      Everitt, Communication Engineering,
      1937, page 439

The Black, French and British patents were not cited by the Patent Office during the prosecution of the patents in suit.

In order to assist the trial court in identifying the elements of the circuits under consideration, a master diagrammatic scheme of the format circuitry was agreed upon and furnished to the court. It was included in the briefs on appeal. Likewise, the format circuits of the accused devices, the patents in suit and the prior art were agreed upon, with certain exceptions, and furnished to us for comparative purposes. The prior art patents are before us.

Plaintiff Becker testified on his own behalf and presented as his expert witness Dr. William L. Everitt, Dean of the School of Engineering at the University of Illinois. Webcor introduced as its expert witness Dr. John A. M. Lyon, Professor of Electrical Engineering at the University of Michigan. Webcor also introduced the testimony of Oscar A. Pearson, its vice-president in charge of research and development, and that of Dr. Richard K. Brown, Professor of Electrical Engineering at the University of Michigan.

Shortly before trial Becker shipped his sole remaining model of the patent circuit outside the jurisdiction of the court. The patents in suit give no values for the circuit components, and Becker could not remember the values he used in the models built by him. Becker made no tests for the court on the circuits disclosed in his patents.

The only tests Becker made were upon accused circuits of Webcor which were shown to be different in certain respects from the Becker circuits. A series of interparty tests were made by the experts employing the test technique referred to as the "Everitt shunt." During the trial Webcor's expert made certain ex parte tests using a modified test procedure, the "Lyon Compensating Impedance."

At the trial, the two opposing experts testified at great length on the procedures used in the various tests, the results obtained and their relationship to the Becker and accused circuits. Lyon testified as to the circuits shown in the prior art. Each expert is admittedly qualified in the field of electronics and demonstrated a high degree of proficiency in the area surrounding the circuits in question. Many of their findings and conclusions were opposite in character, and each generally tended to support the contentions of his respective client.

Everitt concluded that the accused circuits were almost identical to the Becker circuits and that the only differences were minor in character and did not affect the essential operation of the circuits. However, Everitt had never actually tested the Becker circuits and was not later recalled to testify concerning the results of the Lyon ex parte tests, although time was given to afford opportunity for such rebuttal. Everitt did not testify concerning the prior art, having made no tests relative thereto.

In testifying with reference to the Becker circuit and comparing it with a non-Becker connection, Dr. Everitt was frank enough to say:

"I say that I think to understand all the things that are going on in the components of this circuit would require a real study of the order of a Master's thesis, as someone who could put several months' time into completely trying to analyze these capacitants and the other components."

Webcor's expert, Dr. Lyon, found that the results of the Everitt shunt tests made upon the accused circuits were "inconclusive." He concluded from the results of the Lyon modification tests that the alleged push-pull effect claimed by Becker was only negative feedback taught by the prior art. He found the accused devices to be substantially identical to circuits disclosed and taught in the prior art in construction, performance and results attained. He explained that this was caused by negative feedback well known in the prior art. He testified at length concerning the prior art and the teachings to be found there. Certain of his tests related to prior art patents.

We have read and reread the record in this case and have examined the various charts and documentary evidence before us. It is not surprising that counsel and the experts disagree on the meaning and the interpretations to be drawn therefrom. The problem before us relates to a highly technical field of theoretical electronics far beyond the common knowledge of any layman uneducated in this area. We are faced with a situation on appeal here in which we must give great weight to the findings made by a trial judge who saw and heard the experts testify. He had personal opportunity to understand the explanations of the tests, the charts and results produced and the effect to be accorded the extended examination of the witnesses before him.

After the conclusion of the testimony the trial court had the further benefit of briefs by the parties. The court, after indicating a finding favorable to Webcor on all issues, directed Webcor's counsel to prepare and submit proposed findings and conclusions leading to judgment for defendant. This was done. The trial court adopted 38 findings of fact and 18 conclusions of law in the form as submitted. We have carefully examined all of them.

In general, on the disputed issues, the trial court found and concluded that Webcor's accused devices are substantially different from Becker's patented structures; that Webcor's accused devices are substantially identical to circuits disclosed and taught in the prior art; that Webcor's accused circuits do not infringe the patents in suit; and that Becker's patents are invalid when read upon the disclosures of the prior art.

Becker violently objects to such material findings and conclusions as being contrary to and not supported by the record or as being clearly erroneous.

On the issue of infringement Becker had the burden of proof. The question of infringement is one of fact. Hi-Lo TV Antenna Corp. v. Rogers, 7 Cir., 1960, 274 F.2d 661, 665; Kim Bros. v. Hagler, 9 Cir., 1960, 276 F.2d 259, 261-262; Henry Hanger & Display Fixture Corp. of America v. Sel-O-Rak Corp., 5 Cir., 1959, 270 F.2d 635, 641 (design patent); Fritz W. Glitsch & Sons v. Wyatt Metal & Boiler Works, 5 Cir., 1955, 224 F.2d 331, 335; Deller, 3 Walker on Patents, p. 1745.

Becker complains of the repeated findings that he did not test his own devices but only tested the accused structures. The trial court did not hold as a matter of law that one charging infringement is required to test his own devices in order to recover. And, we know of no such strict requirement. However, it is clear that an infringement plaintiff who voluntarily elects not to test his own inventions but chooses only to test the accused devices and rely thereon for recovery may be selecting a tortuous route to travel.

In the case at bar, Becker never convincingly explained why he removed his one remaining model from the jurisdiction of the court and never tested it for the benefit of the court. We think the findings complained of in this respect only go to the point that Becker failed to discharge his burden of proof on the question of infringement. Our examination of the record compels the conclusion that Becker did not sustain his burden of proof of infringement and that the findings of the trial court to that effect are not clearly erroneous. Rule 52(a) of the Rules of Civil Procedure, 28 U.S.C.A.

Based on our examination of the record as a whole, we hold that the trial court properly found that the accused Webcor circuits were substantially identical to the teachings of the prior art in construction, performance and results attained. The findings generally to this effect are not clearly erroneous and must be sustained.

There is support in the record for the findings that the accused structures are substantially different in certain material respects to the patented circuits. Becker complains that there is nothing in the record to warrant such a finding. We do not agree. However, Becker may find himself embarked on an impossible course by insisting that Webcor made "Chinese copies" of his circuits and that they are identical. Since Becker was unable to overcome the demonstration that Webcor's circuits are substantially identical to the prior art, if there is no material difference between the two sets in issue, then it follows that Becker's structures likewise follow the prior art and must be held invalid and not infringed. In any event, we hold that the findings generally of substantial differences between the accused and patented structures are not clearly erroneous.

Finally, we affirm the findings and holding of the trial court that the Becker patents are invalid in view of the disclosures of the prior art. The teachings of the prior art may be learned from the documents before us and from the testimony of Dr. Lyon. We have carefully considered the numerous objections offered against such findings and the many cases cited in support of various standards urged upon us. However, whether we apply the documentary rule in Kiwi Coders Corporation v. Acro Tool & Die Works, 7 Cir., 1957, 250 F.2d 562, or the general rule laid down in Armour & Co. v. Wilson & Co., 7 Cir., 1960, 274 F.2d 143, we come to the same conclusion. We have reviewed the trial court's findings on this factual issue and find that they are not clearly erroneous. In this same category falls the finding to the general effect that Becker made no appreciable improvement over the prior art.

Having found the showing made by the prior art and that Becker has made no substantial improvement over it, the patents are anticipated under 35 U.S.C.A. § 102; and we do not reach the factual question of novelty and the resulting legal question of patentability arising under Section 103 of the Patent Act. Armour & Co. v. Wilson & Co., supra, at pages 156-157.

We agree with Becker that his patents are presumed to be valid, but such presumption has been sufficiently overcome by the results of the tests made herein.

We have considered Becker's complaint that the trial court attached too much importance to the lack of commercial success. It is true that the trial court, in a colloquy with counsel, probably went too far in some of its statements in this connection. However, this was before it finally entered its findings and conclusions which we do not find to be clearly erroneous in that respect.

We have taken into account the challenged finding of invalidity because the patent claims fail to describe the invention in clear, concise and exact terms. While we entertain serious doubt that such a finding has adequate support in the record, we do not find it necessary to pass on that question because of our holding of invalidity on other grounds. The same observation may be made with reference to the finding that the patents in suit failed to assign values to various circuit components.

We have carefully reviewed the other claims of error raised and argued by Becker and find them without merit or that they do not have sufficient relevancy to overcome the holdings announced herein on other critical issues.

We note in conclusion that the parties fail to agree on anything of substance in this case. They do not even concur on whether certain evidence is uncontradicted; on the meaning and effect of much of the expert testimony; on the nature and extent of the tests performed, including those relating to the prior art; or concerning the interpretation fairly to be drawn from these things. We have, however, the firm conviction that the trial court reached the right result and that there are sufficient valid findings to support this result.

The judgment of the district court holding that the claims of the Becker patents under consideration are invalid and not infringed is

Affirmed.