402 F.2d 125
159 U.S.P.Q. 129
UNIVERSITY OF ILLINOIS FOUNDATION, Appellant,
WINEGARD COMPANY, Appellee.
United States Court of Appeals Eighth Circuit.
Sept. 30, 1968, Rehearing Denied Nov. 5, 1968, Certiorari
Denied March 24, 1969, See 89 S.Ct. 1191.
Charles J. Merriam and Basil P. Mann, of Merriam, Marshall, Shapiro & Klose, Chicago, Ill., for appellant; William A. Marshall, Chicago, Ill., and Cook, Blair, Balluff & Nagle, Davenport, Iowa, on the briefs.
Keith J. Kulie, of Burmeister, Kulie, Southard & Godula, Chicago, Ill., for appellee; Donald B. Southard, of Burmeister, Kulie, Southard & Godula, and Edward Dailey of Dailey, Dailey, ruther & Bauer, Burlington, Iowa, on the brief.
Before MATTHES, MEHAFFY and LAY, Circuit Judges.
LAY, Circuit Judge.
The plaintiff, the University of Illinois Foundation, appeals from a judgment below denying patent validity to its United States Letters Patent No. 3,210,767,1 relating to a 'frequency independent unidirectional antenna.' The plaintiff is the owner by assignment from one Dwight E. Isbell. For facility of discussion we refer to the patent itself as the 'Isbell Patent.' Suit was brought against the defendant Winegard Company for alleged infringement. Trial was held before the Honorable Roy L. Stephenson, Chief Judge of the Southern District of Iowa. Judge Stephenson held that the subject matter of the patent did not rise to the level of patentability and dismissed plaintiff's suit. See University of Illinois Foundation v. Winegard Co., 271 F.Supp. 412 (S.D.Iowa 1967).
The Isbell Patent claims a high quality television antenna for color reception, with unidirectional performance over a wide bandwidth of frequencies. The trial court denied validity to the patent in that 'it would have been obvious to one ordinarily skilled in the art and wishing to design a frequency independent unidirectional antenna to combine * * * prior art references * * *.' Id. at 419.
There is no necessity to set forth the full discussion of the facts relating to the prior art. These appear in the excellent analysis made by the Chief Judge below. We have examined the record and find that all claims must be denied, lacking nonobviousness as a matter of law for essentially the same reasons set forth by the court below.
Plaintiff urges that the trial court erred in denying patent validity and calls our attention to a subsequent decision in a federal district court in Illinois which upholds the validity of the same patent against another defendant, University of Illinois Foundation v. Blonder-Tongue Laboratories, Inc., N.D.Ill., Civ. No. 66 C 567. However, as recognized by the district court in Illinois, we must determine this case on its own record.2
The plaintiff basically urges on appeal that the district court in the present case erred in finding that 'those skilled in the art at the time of the Isbell application knew * * * the log periodic method of designing frequency independent antennas.' On the face of the record we must agree that this finding is erroneous. The record discloses that the log-periodic formula3 for antenna design was a recognized theory within the calling. This formula is discussed in the authoritative article by Drs. Hamel and Ore entitled Logarithmically Periodic Antenna Designs, published in March of 1958. However, as pointed out by plaintiff, this same article discloses:
'Unfortunately, no theory has been established which even predicts the types of structures which will give frequency independent operation.'
Jasik's Antenna Engineering Handbook, a recognized authority in the field, verifies on page 18: '* * * it is not possible to determine a priori the frequency independent type of log-periodic antennas.' The record is clear that both plaintiff and defendant experts agreed with this fact.
Thus, we feel the evidence only discloses the adaptation by plaintiff of the log-periodic formula to obtain geometric proportional spacing across its bandwidth. It did not teach that said spacing will in itself achieve frequency independent operation. As we view the record it is the achievement of frequency independent operation through the use of the dipole array which is the principal claim of the Isbell Patent. Both parties' experts agree that with frequency independent operation 'severe end-effects' and unstable performance is avoided. A frequency independent antenna is one in which the basic electrical characteristics remain substantially constant over a given frequency range. These characteristics generally relate to uniform gain, bandwidth, directivity and a good impedance match. Unless an antenna achieves frequency independent operation it will not offer the high quality performance over a wide band of frequencies needed for commercial operation.
However, we do not view the acknowledged fact of the unpredictability of the frequency independent operation of the Isbell Patent as requiring us to reach a result different than held below. Nor was the trial court's finding in this regard essential to its overall reasoning that the patent itself was made obvious by the prior art. The statutory standard of patentability under 103 is not 'predictability.' This terminology helps to obscure the true issue of 'nonobviousness.' The valid test relates the subject patent to whether its development would be obvious to one skilled in the art. As stated by Mr. Justice Clark, the test of nonobviousness 'is one of inquiry, not quality * * *.' Graham v. John Deere Co., 383 U.S. 1, 17, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966). As we have often repeated such inquiry relates to: (1) scope and content of the prior art, (2) differences between the subject patent and the prior art, (3) the level of the skill in the art at the time involved and (4) secondary considerations relating to commercial success, long-felt need, et al. And of course, the latter factors cannot alone determine patentability. National Connector Corp. v. Malco Mfg. Co., 392 F.2d 766 (8 Cir. 1968).
Therefore, we deem the sole legal issue in the instant case to be whether a person skilled in the calling could improve with the skill of the calling the prior art by obvious means.
The mere fact that the 'invention' was not previously known or developed is not the test. Where logical exploration within known principles of the science achieves an unpredictable result, even though a commercially desirable one, the burden of nonobviousness is not necessarily overcome. In other words, the necessity of logical experimentation does not in and of itself negate obviousness.4 Cf. General Mills, Inc. v. Pillsbury Co., 378 F.2d 666 (8 Cir. 1967).
As early asserted by Mr. Justice Bradley in Atlantic Works v. Brady, 107 U.S. 192, 199-200, 2 S.Ct. 225, 231, 27 L.Ed. 438 (1882):
'The process of development in manufactures creates a constant demand for new appliances, which the skill of ordinary head-workmen and engineers is generally adequate to devise, and which, indeed, are the natural and proper outgrowth of such development. Each step forward prepares the way for the next, and each is usually taken by spontaneous trials and attempts in a hundred different places. To grant to a single party a monopoly of every slight advance made, except where the exercise of invention, somewhat above ordinary mechanical or engineering skill, is distinctly shown, is unjust in principle and injurious in its consequences.'
Plaintiff urges Isbell had to overcome two serious handicaps which hid the developmental obviousness: (1) the use of linear dipoles to achieve a broad-band result, and (2) the mismatch of impedance. The simple answer to this argument is that the 'problems' presented were never recognized beyond experimental correction by those skilled in the calling. Although linear dipoles may have been used with 'narrow-band' antennas, such configuration had been shown to be used with antennas having broad-band characteristics. Defendant's tests demonstrated they had substantially the same characteristics (except impedance) as the 'folded dipole.' They had been utilized interchangeably by experimental engineers. These equivalent characteristics were known long before the Isbell Patent.
The problem of matching impedance was one of adjustment. However, matching impedance is an elementary problem to those skilled in radio electronics. Impedance is recognized as the resistance to the flow of current. It is fundamental that efficient performance of any antenna occurs only through a maximum power transfer of the transmission line to the antenna itself. This requires matching impedance as much as possible. No claim of 'invention' is asserted to overcome the mismatch, only that the antenna itself was, therefore, not obvious. The ability to match impedance with the linear dipole array was not new to the development and exploration which led to the Isbell antenna. There was not involved the same deterrent as experienced in United States v. Adams, 383 U.S. 39, 86 S.Ct. 708, 15 L.Ed.2d 572 (1966), where chemicals were interchanged, admittedly non equivalents, to achieve the 'water battery.' The problem here was one of trial and error with a combination of commonly used elements operating within known principles of electronics and mechanics to achieve a desired result. Cf. Marconi Wireless Tel. Co. of America v. United States, 320 U.S. 1, 63 S.Ct. 1393, 87 L.Ed. 1731 (1942).
In Jasik's Antenna Engineering Handbook it is stated on page 1813:
'Since log-periodic antennas are too complex to analyze by present-day theoretical methods, they must be investigated by logical experimental methods.5 However their repetitive nature greatly simplifies the initial experimental investigation because the characteristics need only be measured over one or two periods of frequency.'
Plaintiff claims that the frequency independent operation of the Isbell antenna was accomplished by 'specified scaling factors to determine dipole length, design cell dimensions, and spacing between dipoles.' Here again, the evidence reflects that plaintiff did nothing more than logical and mechanical experimentation within suggested principles and design from the existing art. As the trial court found, there exists a direct relationship between the length and cell dimension of the Isbell claims and those of the prior art found within the K.O. antenna. And plaintiff's own expert discounted the significance of the dipole spacing as a design factor.
Problems of impedance, patterns, resonance, gain, directivity and bandwidth, may all vary upon experimentation within recognized principles of elements having known interrelated electrical characteristics. And naturally, only high quality performance will be the optimum design array desirable. Such experimental combination of common elements and principles, e.g., dipole array, transposed feed, end fed and log-periodic formula, by selected spacing to obtain an improvement of both antenna design and performance, does not obviate the test of nonobviousness. If this were not true, at least in the instant case, nonobviousness would depend simply upon achieving a new result. The standand of patentability under 103 requires more.
Plaintiff's assignor filed his application for patent on May 3, 1960
Compare this court's holding in Imperial Stone Cutters, Inc. v. Schwartz, 370 F.2d 425 (8 Cir. 1966), invalidating the letters patent previously upheld in 1958 by another panel of our court on a different record in Ezee Stone Cutter Mfg. Co. v. Southwest Indus. Prods., Inc., 262 F.2d 183 (8 Cir. 1958)
As summarized by the trial court:
'Generally stated, log periodic antennas are designed according to the theory that an antenna 'design cell' having high performance characteristics for reception of a limited band or period of radio frequency signals, if altered in all dimensions by a constant scale factor will have high performance characteristics for reception of a band of signals having wavelengths which vary from the wavelengths of the first band of frequencies by the same constant scale factor. Thus, according to the theory, if an antenna design cell has certain characteristics for reception of particular frequency wavelengths, an antenna geometrically similar but reduced in all dimensions by a scale factor of .5 will have similar characteristics for reception of frequencies of wave lengths half those of the first.' University of Illinois Foundation v. Winegard Co., 271 F.Supp. 412, 417 (S.D.Iowa 1967).
In Minnesota Mining & Mfg. Co. v. Coe, 69 App.D.C. 217, 99 F.2d 986, 989 (1938), the court said:
'But a showing of great industry in experimental research is not in itself sufficient to constitute invention, when the product thereof differs from those of the prior art only in degree and the result-- no matter how useful it may be-- is merely one step forward in a gradual process of experimentation.'
In Busell Trimmer Co. v. Stevens, 137 U.S. 423, 435, 11 S.Ct. 150, 154, 34 L.Ed. 719 (1890), the United States Supreme Court said:
'But the patent before us is no such case. The most that can be said of it is that it shows, on the part of Orcutt, great industry in acquiring a thorough knowledge of what others had done in the attempt to trim shoe soles in a rapid and improved mode, by the various devices perfected by patents for that purpose, good judgment in selecting and combining the best of them, with no little mechanical skill in their application; but it presents no discoverable trace of the exercise of original thought.'
Mr. Winegard of the defendant company relates the experimental engineers' approach:
'Well, in the early days I used to calculate the resonance. There's so many factors with respect to the length of the element-- if you are just concerned with one dipole, one straight dipole, it would be very easy to determine the actual length by calculation, but when you place several elements together there is always an inter-action which will affect the length, so in order to get optimum performance from an array it is necessary to tune the antenna by slightly modifying the actual length of the-- I have been working with antennas so long it is like learning the English language, you don't have to look at the dictionary every time to figure out how to spell 'and.' You know what I mean?'