271 F.2d 218
122 U.S.P.Q. 180
HAZELTINE RESEARCH, INC., Plaintiff-Appellant,
DAGE ELECTRIC COMPANY, INC., Defendant-Appellee.
United States Court of Appeals Seventh Circuit.
July 8, 1959, Petition for Rehearing Withdrawn Nov. 16, 1959.
Edward A. Ruestow, Little Neck, N.Y., M. Hudson Rathburn, Chicago, Ill., W. Houston Kenyon, Jr., New York City, Leroy Eason, Little Neck, N.Y., Mason, Kolehmainen, Rathburn & Wyss, Chicago, Ill., of counsel, for appellant.
Theodore W. Anderson, Jr., Chicago, Ill., C. B. Dutton, Indianapolis, Ind., Olson, Mecklenburger, von Holst, Pendleton & Neuman, Chicago, Ill., of counsel, for appellee.
Before SCHNACKENBERG, PARKINSON and KNOCH, Circuit Judges.
PARKINSON, Circuit Judge.
Plaintiff-appellant, Hazeltine Research, Inc. (Hazeltine), brought this action for infringement of its Patent No. 2,540,012 against Dage Electric Company, Inc. (Dage), defendant-appellee. Dage counterclaimed for a declaratory judgment of invalidity and noninfringement. The District Court found the patent invalid and dismissed the complaint, but also found the patent, if valid, to have been infringed. This appeal followed.
The patent in suit was issued, on January 30, 1951, to Hazeltine, as assignee of one Octavio M. Salati, the inventor, on application of Salati filed May 19, 1945.
The patent in suit deals with a connector used in the joining together of coaxial transmission lines or cables. These cables, used extensively in the electronics industry, carry relatively high frequency electrical energy. The high frequency necessitates the use of a specialized transmission line and as 'coaxial' implies there are, in the cable, two conductors, as inner rod and an outer tubular sheath, separated by an insulating dielectric spacer. The cable itself, if properly designed, creates no difficulty and its ability to transmit the electrical energy with a minimum of reflection back towards the source can be adequately determined. The engineering term 'characteristic impedance' is one used to denote the nonreflection factor of any coaxial cable, i.e., if the characteristic impedance is constant there will be no reflection.
The difficulty arises when one coaxial cable must be joined to another. If the cable connector is such that the characteristic impedance changes abruptly there will be a reflection. Such a change is known as a 'discontinuity'. There is also the further problem of a short circuit between the inner and outer conductors of the connector if not properly designed as to the dielectric spacer.
Without going into greater detail it appears that Hazeltine admits that the discontinuity problem presented by the connectors had been solved insofar as medium and large coaxial cables were concerned. However, it contends that its patented connector was and still is the only adequate one in the field of small size cables, or, as it says, the problem was 'miniaturization', i.e., how to solve the reflection problem in small size connectors while still meeting the problems of voltage breakdown and sturdy mechanical construction.
Hazeltine argues that the prior art connectors were of the 'continuous constant impedance' type in which the impedance is the same for any portion, however short, along the connector length. These connectors were designed so that their impedance 'matches' the impedance of the cable with which they are to be used. However, if this type is utilized to connect cables of a relatively small size either the connector would be of a larger physical size than desirable or if made in a small size the components, particularly the inner rod would be fragile, the tolerances would become difficult to maintain, and such connectors would not be suited to mass production. This is due to the fact that there is a set ratio between the inside radius of the outer sheath and the radius of the inner rod and as the former becomes smaller so must the latter, hence the size of the connector is limited to the minimum practical size of the inner rod. Of course, to ignore the ratio between the two would not give the desired impedance. Hazeltine, however, contends that Salati overcame these obstacles simply by abandoning the continuous constant impedance theory and substituting therefor sections of impedance in his connector that were either higher (inductive) or lower (capacitive) than line impedance and which averaged out to line impedance.
Using the Salati principle it is possible to construct a connector or small size wherein the internal components have the necessary size, strength and rigidity required for field use inasmuch as the prior ratio between the inner rod and outer sheath need no longer be maintained through each incremental length of the connector. However, Dage contends that the prior art actually set forth the same principles that Salati claims for his invention and that Salati's patent is vague and indefinite.
The District Court made extensive findings of fact. Basically they may be broken down into the four following main findings:
1. The language of the claims of the patent in suit is invalid in that it is vague and indefinite;
2. The patent taught nothing not already found in the prior patents relied upon by Dage;
3. The patent taught nothing not already found in the three connectors designed for the Armed Forces which constituted prior knowledge; and
4. If the patent in suit is valid it has been infringed by the two Dage accused devices.
The District Court based its findings of vagueness and indefiniteness on the following:
1. It is silent as to size, type of characteristic impedance of the cable with which it is to be used, and it provides no information from which these facts might be determined;
2. It is silent as to the size of the connector or what would constitute a small connector;
3. It uses the term 'so proportioned' when actually three definite proportions as to dimensions and material would be necessary; and
4. It fails to mention the existence of discontinuity capacitance, the method of determination, or the inclusion of the factor in the over all design.
Assuming that these findings are correct in their factual content, do they then invalidate Salati's patent as being vague and indefinite? We think not.
It is well settled that a patent must be written in a manner intelligible to the person who would normally be expected to utilize it. The fact that a patent would use such terms or, to the converse, omit such terms as would then render it incomprehensible to a layman would not of necessity render it invalid. It is only when those skilled in the art are unable to determine the extent, purpose or rationale of the invention from a study of the patent that it may be said the patent is vague and indefinite and, therefore, invalid.
The Supreme Court has repeatedly held that a specification in letters patent is sufficiently clear and descriptive when expressed in terms intelligible to a person skilled in the art to which it relates. Webster Loom Co. v. Higgins, 1881, 105 U.S. 580, 26 L.Ed. 1177; Seabury v. Am Ende., 1894, 152 U.S. 561, 567, 14 S.Ct. 683, 38 L.Ed. 553; Carnegie Steel Co. v. Cambria Iron Co., 1902, 185 U.S. 403, 437, 22 S.Ct. 698, 46 L.Ed. 968.
Therefore, determination of the question of vagueness and uncertainty must be made in light of what a person skilled in the electronics are would be able to glean from the Salati patent.
The District Court made, inter alia, the following findings of fact:
'Coaxial cables and connectors therefor are old and well known, dating back prior to 1930. The transmission and treatment of microwaves was well understood prior to 1900.
'A. As early as 1924 the textbooks included sections teaching the design of coaxial transmission lines including determination of capacitance, C, inductance, L, and characteristic impedance Zo from the ratio of the conductor diameters D/d and the dielectric constant, K. This teaching was utilized in the design of cables, connectors or other associated appurtenances.
'B. The attainment of high voltage operation by providing long surface paths between the inner and outer conductors was also well known as early as 1930, and tables, charts and formulas of voltage breakdown through air were available.
'C. During 1944 two articles (Defendant's Exhibit 26) by Whinnery et al. appeared in the Proceedings of the Institute of Radio Engineers teaching the importance of and design compensation for discontinuity capacitances which are the important reflective effects produced at each radial wall or step within a coaxial cable or connector.'
It would appear that the full significance of these findings was not correlated by the District Court to the question of the vagueness of the patent in suit. All that was necessary for Salati to set forth was the method of designing the well understood components of the connector so that there was an enlarged inner rod, an adequate dielectric spacer and compensating sections of high and low impedance. The term 'so proportioned' simply referred the person reading the patent to the formulae, of which many were known since 1924, in order that he might design his connector with an average impedance matching that of the cable to be used. The fact that Dage's expert witness was able to analyze the Salati patent indicates that those skilled in the art would experience no difficulty in comprehending its teachings. Hazeltine's expert witness was asked the question: 'Would you say whether or not in your opinion the disclosure in the Salati Patent as a whole is adequate to teach one skilled in the art how to practice this Salati invention?' His answer was: 'Yes, I would think so.' We find no evidence in this record to indicate Salati's patent would be unintelligible to those skilled in the art. We think under the circumstances the term 'so proportioned' was adequate for those who would make use of Salati's patent. Webster Loom Co. v. Higgins, 1881, 105 U.S. 580, 26 L.Ed. 1177.
The problem as to whether Salati's patent is invalid in that it does not state the size or impedance of the cable with which it is to be used nor what constitutes a 'small connector' is also illusory. While it is true that the Salati connector was designed originally for 1/4 inch 50-ohm coaxial cable and that its greatest utility lies in the smaller sizes where the prior type connectors are unsatisfactory, a person skilled in the art, reading the Salati patent, would realize that the Salati connector could be employed with either a small or large cable, providing, of course, the proper dimensions as found in the proven formulae were used. The fact the patent does not specify an exact size does not render it invalid.
The failure to mention the possibility of a discontinuity capacitance occurring within Salati's connector and the significance thereof is, in our opinion, not an invalidating defect. A discontinuity capacitance caused by any sharp curve in the line or, in the case of a connector, any acute break on the surface of the conductors, may cause a reflection. In the year 1944 the so- called 'Whinnery articles' were published whereby it was possible to compute the discontinuity capacitance of any irregularity that might occur within a connector. These articles predated Salati's application. Therefore, both the presence and significance of discontinuity capacitances were known to those skilled in the art and would simply have to be taken into account when making use of the patent, i.e., when computing the dimensions of the inductive section to balance the capacitive section, the discontinuity capacitance must also be considered.
It might appear that the term 'so proportioned' is being allowed too much latitude; however, a careful review of the record convinces us that a person skilled in the art would have no difficulty in following Salati's teachings and factors such as the discontinuity capacitance would present no problem to those who keep abreast of the latest findings in the field of electronics.
We, therefore, hold that the Salati patent is not invalid for want of specificity.
The District Court found that Salati's patent taught nothing not already found in the prior art. Nine separate patents were cited by Dage.
The first four patents, Green No. 1,841,473; White, et al. No. 2,125,597; Kaar No. 2,127,408; and Fell No. 2,406,945. were used, as stated by Dage, 'to establish that the patent art bore out the equations, sample calculations, and extensive connector design background provided by Dean Fischer at the trial.' These patents simply constitute the general background of the art and buttress our holding that one skilled in the art would not find Salati's patent vague or indefinite.
The White and Kaar patents are more or less the same; both deal with the problem of joining together cables of two different impedances. There is uncontradicted testimony by Hazeltine's expert that neither White nor Kaar attempted to solve the problem that Salati faced nor does their teaching in any way direct itself to a solution of that problem. The Green cable joint is to be employed with high frequency cables wherein the impedances are the same but the physical sizes differ between the two cables to be joined. Fell's patent is an 'insulator for concentric transmission lines.' None of these patents deals with an easily detachable connector.
We now consider the remaining five patents to determine if, individually or in combination, they had brought the art to such a point that Salati's teachings were either known or were obvious as the next logical step to those skilled in the art.
Watts No. 1,871,397, as well as Waite, Jr. No. 2,615,953, is primarily concerned with the mechanics of the connector. Both seem to be aware of the problem of discontinuity but other than casual reference to avoid it there is nothing. As to Waite, Dean Fischer, Dage's expert witness, testified, 'the Waite patent really is directed towards the mechanism or device of clamping means and the way in which the braid is claimped and the way in which the insulation of the cable can also be clamped at the end of the cable.'
Buschbeck No. 2,267,371 provides a compensating section in a cable where a joint, connector or sharp bend has produced a discontinuity that must be cancelled out. However, hazeltine's expert testified, and we can find nothing in the record to the contrary, that if Buschbeck's section was reduced to the physical size that a Salati connector might be, the 'intermediate conductor * * * would become impossibly fragile. It would not be a rigid member anymore.' Also in the Buschbeck patent the two capacitive sections and the one inductive section were arranged in a manner contrary to the sequence Salati used. On its face this distinction might not appear too important but again Hazeltine's expert, without contradiction, testified that for Salati to have followed Buschbeck would have 'spoiled the other features of his arrangement', i.e., the size and spacing of the dielectric. In short, the only testimony anent the direct comparison of Buschbeck to Salati, as opposed to Dean Fischer's discussion of Buschbeck standing alone, was to the effect that Salati never could have made a workable miniaturized connector had he followed the teachings of Buschbeck.
Cork No. 2,490,622 is basically a constant continuous impedance connector with a screw and ring adjustment to from a complete impedance match. The screw provides a capacitive section as it is threaded in or out as desired. Also by means of an adjusting ring the insulation material of the halves may be kept at a desired distance apart thereby forming an air space which would constitute an inductive section. Hence, as Dage says, this connector might have both an inductive and capacitive section. However, a comparison of the two patents shows that the teachings of Cork would not lead to the patent in suit.
Cork's patent utilizes two devices known to the prior art. The inner and outer conductors are kept a uniform distance apart thereby giving a constant impedance and the connector itself is one-half wave length long. Hence if this connector was miniaturized Cork would simply be faced with the problems always attendant on the smaller sized constant impedance connectors. Moreover, Cork's inductive and capacitive sections are mere corrective devices created in a totally different way than those of Salati. Cork's device would not be satisfactory in field use as would Salati's in that the former would have two moving parts on the exterior surface.
The remaining patent is Weber, et al. No. 2,529,436 which shows a 'modified form of bullet connector.' The patent goes on to say that the 'purpose of providing the groove 5c is to compensate for the discontinuity in the conductor shape used by the collar 5b, in other words, any wave reflection caused by collar 5b is compensated or cancelled by a corresponding reflection caused by groove 5c'. Here Weber has apparently a normal connector wherein, due to a sharp break caused by the attachment of the 'bullet' to the inner rod, a discontinuity appears. This solved by a 'groove', i.e., the adjacent diameter of the 'bullet' is sharply reduced a corresponding amount and, therefore, the discontinuity is cancelled out. Here again we have basically the constant impedance connector, with all the inherent problems of miniaturization, with a corrective factor added. Without going into greater detail we can say that while Weber, to some extent, actually used a compensating section in a prior type connector to cancel a discontinuity the teachings are such that Weber would not have made obvious, to a person skilled in the art, a connector where such compensating sections constituted the very essence thereof.
As to the three nonpatented connectors, HN, BN and BNC, developed for the Armed Forces, constituting prior knowledge, the District Court found that the 'complete combination of even the narrowest and most detailed claims of the patent is completely shown in * * * the BN, HN and Bell BNC connectors.'
The type HN is a matched connector, i.e., it has the same impedance as the cable to which it is attached. There is no testimony in the record that this connector used the mismatching technique of Salati's patent but to the contrary it appears that it was of a large size and was a continuous constant impedance connector of the prior style. Despite Dage's contentions in its brief, Dean Fischer did not testify as to whether the HN connector had sections of high or low impedance and in fact said nothing other than the HN was 'said to be a constant impedance connector.' The record is barren of any substantial proof that the HN connector resembles in any way the Salati connector or taught anything that would have aided Salati.
The type BN connector, while having successively different impedances, was not intended for high frequency transmission and at such high frequency would be completely unsatisfactory. The Government publication Installation and Maintenance of Transmission Lines Waveguides and Fittings issued by the Electronics Division of the Bureau of Ships, Navy Department, states as to the BN connectors: 'They are not constant impedance and are therefore not recommended for applications with frequencies in excess of approximately 200 megacycles per second unless the electrical requirements of the circuit are not critical in regard to impedance matching,'
We fail to see how this connector could have taught Salati anything about high frequency connectors.
Lastly we come to the BNC connector upon which Dage relies heavily. We have searched the record and we can find no testimony indicating that the BNC used Salati's mismatching technique. Dage's expert witness, Dean Fischer, testified that the BNC would be matched only with a cable having approximately 52 ohms. This in itself means nothing as the matching may be accomplished in many ways. Hazeltine's expert and Salati both testified that the BNC would be badly mismatched and was completely unsatisfactory because of excessive reflections.
On September 19, 1944 Lt. Cdr. Neher, identified as 'chariman Army-Navy R. F. Cable Coordinating Committee', wrote to Hazeltine in part as follows:
'At the last meeting of members of the Army-Navy R. F. Cable Co-ordinating Committee and your organization to discuss constant impedance connectors for RG-58/U cable, the design presented by your organization looked very good and sufficiently near completion to be considered the standard type for its intended application.
'In accordance with the information obtained at this meeting, the Coordinating Committee discontinued their own work on this type connector and recommended that Bell Telephone Laboratories discontinue their work.'
Dage's own witness, Edwin A. Mroz, a Governent Electronic Engineer, testified that he suspected that it was the BNC connector which was recommended for abandonment.
Therefore, we must conclude that there is no valid basis upon which the District Court could find that the HN, BN or BNC connectors constituted invalidating prior knowledge. The only supported finding from this record could be that the Salati patent superseded and, to a great extent, made obsolete the three aforementioned connectors, and that the patent in suit was not invalid over the prior art. In fact, Hazeltine's expert testified, without objection or contradiction, as follows:
'Q. Tell me, Professor, whether or not in your opinion the Salati Patent would have been obvious to one skilled in the art at the time it was made?
'A. No, sir, in my opinion it certainly would not have been obvious.'
The burden of proving invalidity by clear and cogent evidence was on Dage. Radio Corp. v. Radio Eng. Laboratories, 1934, 293 U.S. 1, 8, 55 S.Ct. 928, 79 L.Ed. 163; Mumm v. Jacob E. Decker & Sons, 1937, 301 U.S. 168, 171, 57 S.Ct. 675, 81 L.Ed. 983; Helms Products v. Lake Shore Manufacturing Co., 7 Cir., 1955, 227 F.2d 677, 680. We have in mind the decisions of this Court in Hobbs and Hyster,1 nevertheless, we have found nothing in the prior art or knowledge which rebuts the prima facie proof of the patent's validity as the Salati patent is valid over the prior art, including that which was not before the Patent Office.
A brief study of the case of Great A & P Tea Co. v. Supermarket Equipment Corp., 1950, 340 U.S. 147, 71 S.Ct. 127, 95 L.Ed. 162, will indicate how difficult it is to set an exact standard of invention where, as here, no new physical element has been added. However, salati went for beyond the simple realignment of old and known components. It is true that the inner conductor, outer conductor and dielectric were all well known to the art but they are by nature necessities for all connectors in this field. Salati not only rearranged these parts but in doing so injected an entirely different concept that theretofore had not been thought of in relation to coaxial cable connectors, and produced an entirely new and useful result. We recognize that compensating sections had been used in the industry from time to time as a corrective factor but that the idea of using this principle for the basis of connector design was not obvious to those well skilled in the art is shown by the lack of success Bell Telephone Laboratories had with the BN and BNC connectors. Rohm & Haas Company v. Roberts Chemicals, 4 Cir., 1957, 245 F.2d 693, 697; Technical Tape Corp. v. Minnesota Mining & Mfg. Co., 2 Cir., 1957, 247 F.2d 343, 347.
Looking backward we may say the principle discovered by Salati was lurking in the background but it is clear from this record it was not obvious to one skilled in the art. Hindsight is of little value now.
The District Court found that two of the accused devices manufactured and sold by Dage infringed the patent in suit, if the patent is valid. Dage does not question that finding here and we agree that Dage is guilty of infringement as to the two accused devices, No. A-100-056-1 (Military Type No. UG 88/U) and No. A-100-130-1 (Military Type No. UG 89/U).
For the reasons stated the findings and conclusions of the District Court that the patent in suit is invalid are not supported by any substantial evidence and are clearly erroneous.
The District Court made other findings with which we cannot agree. However, as they are of minor significance and in light of what we have heretofore held any further discussion would serve no useful purpose.
Judgment reversed and cause remanded for further proceedings consistent with this opinion.