OPINION
ANDEWELT, Judge.In this patent action filed pursuant to 28 U.S.C. § 1498, plaintiff, William A. Casler, seeks compensation from the United States for the alleged unauthorized use of a device allegedly covered by Claims 1, 3, and 4 of United States Patent No. 3,259,727 (the ’727 patent). The device in issue is a roll ring rotary electrical connector manufactured for the United States by Sperry Flight Systems (Sperry), a division of Sperry Corp. Defendant contends that it is not liable under 28 U.S.C. § 1498(a), because (1) plaintiff’s ’727 patent is invalid, and (2) in any event, the ’727 patent does not cover Sperry’s roll ring connector. This case was *719tried before a judge of this court who resigned before rendering a decision. After the case was reassigned, the parties requested that the case be decided on the existing record. For the reasons set forth herein, based on a comprehensive review of the trial record, the court finds that (1) defendant has not satisfied its burden to demonstrate that the ’727 patent is invalid, but (2) plaintiff has failed to demonstrate that the ’727 patent covers the Sperry roll ring device. Judgment therefore is awarded to defendant.
• I. Brief Background — Rotary Electrical Connectors
Electrical connectors are devices used to provide electrical contacts for connecting and/or controlling electrical circuits. Electrical connectors are manufactured in a variety of designs and employ differing technologies, depending upon the environment in which they are used and the particular functions they must serve.
Rotary electrical connectors are a specialized type of electrical connector that is used to pass electricity between conducting elements that are mutually rotatable, i.e., where one or both elements rotate with respect to the other. Rotary connectors are used, for example, to conduct electrical signals between a rotating antenna and a stationary circuitry housed in the base of the antenna, or between the rotating and stationary parts of a gyroscope. Where only one conducting element rotates, the rotating element is generally referred to as the “rotor” and the stationary element as the “stator.”
Maintaining a constant electrical circuit between elements that rotate with respect to each other presents a series of special problems. First, vibration within the apparatus or bumping of the apparatus from the outside can result in momentary lapses in contact and a consequent interruption in the flow of electricity. Next, a condition known as “background noise” can result when there are increases in resistance to the flow of electricity at the surfaces of the rotating elements, e.g., when the surfaces of the elements are subject to oxidation, minute contaminants, or wear. If the level of the increased resistance is sufficiently high relative to the strength of the primary electrical signal, the noise can obscure, if not totally mask, the primary signal. Background noise is especially troublesome when the electrical signal itself is weak, i.e., characterized by low voltage and low current.
Another problem confronted when conducting electricity between mutually rotatable surfaces is that when pressure is used to maintain contact between the conducting elements, a torque can result, i.e., a force that tends to produce rotation. Torque is not a significant problem for many uses, but can be for some, such as when rotary connectors are employed in gyroscopes used to keep satellites on a particular track. Torque can cause the gyroscope to drift and produce unintended changes in satellite direction.
Over the years, a variety of different technologies have been used in the manufacture of rotary electrical connectors. A technology that emerged in the early 1900s involves placing rolling discs or ball bearings between the rotating elements. The discs and ball bearings maintain contact between the mutually rotating conducting elements with electricity passing from one conducting element through the rolling discs or bearings to the other conducting element.
A later-developed technology involves the use of slip rings. A slip ring assembly commonly uses a stationary arm with a metal or composite brush at its end. The arm is attached to or is a part of one conducting element and the brush presses against the other conducting element, typically a rotating drum. Electricity is transferred continuously between the brush and the drum as the brush rubs or slides along the surface of the drum. Because slip ring technology involves rubbing or sliding against the drum surface, a torque is created. In addition, lubricant contaminants sometimes build up between the drum and the surface of the brush which can result in background noise. The build up of contaminants can be mitigated somewhat by *720increasing the pressure between the brush and the surface, but increased pressure also increases coupling torque and mechanical wear.
This litigation involves yet a third technology known as rolling ring technology in which a flexible resilient ring or loop is placed between rotating conducting elements. When one or both conducting elements rotate, the ring rolls around between the elements, maintaining constant contact between them. The original inventor of rolling ring technology is Nils E. Linden-blad, who described the technology in United States Patent No. 2,467,758, issued on April 19, 1949.
II. The 727 Patent
A. Description of the Casler Connector in the Specification of the ’727 Patent
Plaintiff is the sole inventor and sole owner of the ’727 patent. The patent, issued on July 5, 1966, and entitled “Low-Resistance Connector,” covers a particular design of a roll ring rotary electrical connector.1
The specification describes the “principal purpose” of the Casler invention as the provision of a rotary electrical connector that is effective in situations where low, constant resistance circuit paths are required, e.g., where the electrical signal is characterized by low voltage and low current. The connector disclosed therein is made up of three elements — two circular concentric conductive elements, one or both of which are rotatable relative to the other, and an intermediate transfer contact, i.e., the rolling ring, which effectively completes the circuit between the conductive elements.
The relative position of the two conductive elements, also called contact rings, and the position of the transfer contact between the conductive elements can be seen in Figure 2 of the ’727 patent. The inner conductive element is labeled element 9, the outer conductive element is labeled element 17, and the intermediate transfer contact, which is the kidney-shaped spring between the conductive elements, is labeled element 25.
[[Image here]]
*721. Both the conductive elements and the transfer contact are made of electrically conductive materials. The inner and outer conductive elements are positioned in a concentric and coplanar fashion, i.e., the inner element is located inside the outer element and both elements rotate about the same axis of rotation. The inner conductive element has a generally V-shaped groove located along its outer periphery, and the outer conductive element has a generally V-shaped groove located along its inner periphery. The opposing grooves, not visible in Figure 2, are shown in Figure 1, which is an axial sectional view of a two-layer contact. The V-shaped grooves are labeled elements 11 and 19.
[[Image here]]
The rolling ring intermediate transfer contact, also referred to as a flexure, spring, or loop, completes the circuit. The transfer contact, a flexible, highly resilient loop, is placed in the radial gap between the two conductive elements and makes contact with the generally V-shaped grooves of both the inner and outer elements. As shown in Figure 2, when the transfer contact is placed between the conductive elements, it forms an arc around the inner element and is thereby distorted into a generally kidney-shaped form.
When in operation, either one or both of the conductive elements rotate, continuously or intermittently, relative to the other. The intermediate transfer contact rolls in and out of the generally V-shaped grooves of the respective elements, “tracking” along the sides (not the bottom) of the grooves. The radial separation between the contact rings remains fixed, as does the generally kidney-shaped distortion of the transfer contact and the radial force exerted by the transfer contact on the conductive elements. The inherent resiliency of the transfer contact provides the necessary radial force to secure the contact firmly between the elements.
The specification states that this particular design produces an improved electrical connection. First, the design results in stationary, nonsliding, high-unit contact pressure between the intermediate transfer contact and the two rotatable elements that is virtually unaffected by vibration. The high-pressure contact results from the resilient transfer contact being “wedged” into the V-shaped grooves of the conductive elements. The high-contact pressure exists at each point at which the transfer contact touches the groove walls along the arc of contact. Because there is a plurality of such points along the arc of contact, at all times there is a plurality of low-resistance, high-contact pressure parallel paths for the electricity to flow from one conductive element through the transfer contact to the other conductive element.
Second, the specification indicates that the Casler design limits undesired noise by effectively removing dust, oxidation, and wear particles from the conductive elements. In the regions where the transfer contact enters and leaves the contacting *722surfaces of the element grooves, in addition to the wedging action, the individual coils are subjected to a turning action, i.e., a change in radial orientation that causes a scraping or wiping action and cleans the contacting surface of each conductive element.
In addition to the rotary connector described above, the ’727 patent also describes and claims a stationary switch that employs a somewhat similar design.
B. The Claims of the ’727 Patent in Issue
Claims 1, 3, and 4 of the ’727 patent each covers a Casler rotary connector. At trial, the parties disputed the meaning of certain of the claim terms. These disputes are relevant to the issue of the scope of the claims discussed infra. The claims are reproduced below, and the disputed claim terms are underlined.
1. Means for establishing a low resistance electrical connection between mutually rotatable parts comprising an outer contact ring having a generally Y-shaped groove formed in the inner periphery thereof,
a circular contact member mounted substantially coaxially within said contact ring for rotation relative thereto and having a generally V-shaped groove formed in its outer periphery substantially coplanar with and facing the groove in said contact ring, and an endless loop intermediate transfer contact including
a spring of generally toroidal form distorted into a generally kidney shaped form and
wedged between the Y-shaped grooves facing each other,
the cross-section of the spring being less than the radial separation between said ring and said circular member and
the diameter of the undistorted loop being greater than the radial separation between said ring and said circular member so that when so distorted there is an appreciable arc of contact between said spring and each of said ring and said circular contact member.
3. Apparatus as defined in claim 1 wherein
said transfer contact comprises a single endless loop of spring wire.
4. Means for forming electrical contact between mutually movable parts comprising
a pair of contact members mounted in substantially coplanar relationship on said parts respectively,
each of said contact members having a V-shaped groove formed therein so that said grooves face each other and are spaced by a substantially constant distance throughout their extent, and a transfer contact comprising
a continuous spring forming a loop the diameter of which when undistorted is greater than the spacing between said grooves and
a cross-section of which is less than the spacing between said grooves,
said spring being distorted into a generally kidney-shaped form and wedged into each of the facing grooves.
III. Assessment of Validity of the ’727 Patent
A. Legal Standards
Defendant contests the validity of each of Claims 1, 3, and 4 of the ’727 patent, alleging that the claims fail to meet the statutory requirement of nonobviousness. Pursuant to 35 U.S.C. §§ 103 and 282, the claims in a patent are invalid if “the differences between the subject matter [of the claims] 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 one of ordinary skill in the art to which the subject matter pertains.” The claims in an issued patent are presumed valid, 35 U.S.C. § 282, and the party challenging validity has the burden of proving invalidity by clear and convincing evidence. Atlas Powder Co. v. E.I. Du Pont de Nemours & Co., 750 F.2d 1569, 1573, 224 USPQ 409, 411 (Fed.Cir.1984).
*723A determination on the issue of non-obviousness entails a conclusion of law. Environmental Designs, Ltd. v. Union Oil Co. of California, 713 F.2d 693, 695, 218 USPQ 865, 867 (Fed.Cir.1983), cert. denied, 464 U.S. 1043, 104 S.Ct. 709, 79 L.Ed. 2d 173 (1984). However, that legal conclusion must be based upon a series of factual determinations. The court must evaluate (1) the scope and content of the prior art, (2) the differences between the prior art and the claimed invention, (3) the level of ordinary skill in the pertinent art, and (4) evidence directed at so-called secondary considerations, which may serve as indicia of nonobviousness. Graham v. John Deere Co., 383 U.S. 1, 17, 86 S.Ct. 684, 693, 15 L.Ed.2d 545, 148 USPQ 459, 467 (1966); Atlas Powder, 750 F.2d at 1574, 224 USPQ at 412; Environmental Designs, 713 F.2d at 695, 218 USPQ at 867. Herein, the parties agree that the appropriate level of ordinary skill in the pertinent art is a designer of rotating electrical connectors with a college degree in either electrical or mechanical engineering.
B. Scope and Content of the Prior Art and Differences Between the Prior Art and the Claimed Invention
In evaluating the patentability of the invention claimed in the '727 patent, the Patent Office focused primarily on four prior art patents: Lindenblad (United States Patent No. 2,467,758), Boeye (United States Patent No. 2,307,192), Hopper (United States Patent No. 2,449,479), and Linn (United States Patent No. 2,154,275). In contesting the validity of the patent at trial, defendant also relied on the Lindenblad patent, but contended that other prior art, not cited by the Patent Office during the patent prosecution, is more significant than the Hopper, Boeye, and Linn patents.
The trial court insisted that defendant limit its evidence to only the most pertinent prior art. Thereafter, defendant relied on Lindenblad and three other references: United States Patent No. 2,359,055, issued to A.C. Schwager; United States Patent No. 2,788,420, issued to J.A. Mathews; and a 1940 textbook by G. Windred, entitled “Electrical Contacts.”2 This court agrees with defendant that these four references constitute prior art for purposes of evaluating the obviousness of the claims in issue. See In re Deminski, 796 F.2d 436, 441-42, 230 USPQ 313, 315 (Fed.Cir.1986).
The Lindenblad Patent
As indicated above, the Lindenblad patent, entitled “Moveable Circuit Closure,” opened a new chapter in rotary electrical connectors by disclosing the use of a flexible loop to carry electricity between mutually rotatable surfaces. The specification indicates that the Lindenblad connector, designed to transfer high-voltage, high-amperage electrical current through a rotating joint, provided “an improved circuit closure which wears longer and has a wider contact surface area than conventional arrangements.”
The Lindenblad design employs concentrically oriented inner and outer conductive elements and a flexible band transfer contact that assumes a generally elliptical or kidney-shaped form when compressed between the conductive elements. As shown in Figure 4 of the Lindenblad patent, the transfer contact rides against the cylindrical surfaces of the conductive elements, and thereby generates a substantial area of contact between the elements and the transfer contact.
[[Image here]]
*724Unlike the design described in Claims 1, 3, and 4 of the Casler patent, Lindenblad does not employ grooves in the opposing peripheries of the conductive elements. Rather, Lindenblad uses either flanges or pins to keep the transfer contact in place and to achieve constant contact between the conductive elements and the transfer contact. Moreover, as can be seen in Figure 4, because of the absence of generally V-shaped grooves in Lindenblad, the transfer contact touches the entire surface of the conductive elements rather than just the edges of a groove. The larger mating surfaces in Lindenblad result in lower localized pressures between the transfer contact and the conductive elements than are present in the Casler V-shaped groove design.
The Schwager Patent
The Schwager patent, entitled “Roller Contact for Hinged Conductors,” was issued on September 26, 1944, predating Lin-denblad by almost five years. Schwager is also directed at maintaining an electrical connection between relatively movable conductive elements. The Schwager elements, however, apparently do not rotate completely with respect to each other, but rather are hinged and have a limited range of relative motion.
Defendant relies upon Figure 9 of the Schwager patent, which is as follows:
[[Image here]]
Defendant’s expert witness, Peter Jacobson,3 testified that the device depicted in Figure 9 transfers electricity between an inner rotating member with a V-shaped groove and an outer member with a V-shaped groove through use of a transfer contact located between the opposing V-shaped grooves. According to Jacobson, in Figure 9, the V-shaped groove on the inner rotating member is made up of elements 103 and 104, the outer V-shaped groove is made up of elements 107 and 108, and wedged between these grooves is element 105, a ball bearing that acts as an intermediate transfer contact. Electricity is conducted from the inner V-shaped groove to the outer V-shaped groove. In the following diagram, the V-shaped grooves Jacobson found in Figure 9 are darkened and the arrow is inserted to indicate the direction in which he alleged the current flows between these V-shaped grooves.
[[Image here]]
Plaintiff did not dispute Jacobson’s testimony on this point. But after careful consideration of Figure 9 and of the specification of the Schwager patent, this court concludes that Jacobson’s description of the structure and operation of the device is incorrect. In Figure 9, the electricity does not flow from what Jacobson characterizes as the inner V-shaped groove to what he characterizes as the outer V-shaped groove. To the contrary, the functioning electrical contact elements in Figure 9 are the elements labeled 103 and 104, respectively, which Jacobson described as the two sides of an inner V-shaped groove. Thus, in Figure 9, the electricity actually passes from one side of what Jacobson characterizes as the inner V-shaped groove to the other side of that same groove. See Schwager patent at col. 2, lines 68-72. The structure that Jacobson described as an outer V-shaped groove is in fact not a groove in a conductive element, but rather is an adjustable, free-floating outer assem*725bly used to pressure the transfer contact in the direction of the conductive elements. Thus, contrary to Jacobson’s testimony, the Schwager patent, unlike Casler, does not disclose inner and outer conductive elements having generally V-shaped grooves formed in their peripheries.
In any event, under either Jacobson’s or this court’s interpretation of Figure 9, the Schwager device would differ from the subject matter covered in the relevant Cas-ler claims in certain other respects. First, the only embodiments of the generally V-shaped grooves disclosed in the Casler specification are unitary grooves in the general shape of the letter V, with two sides that intersect. The alleged grooves in the Schwager patent, however, are each made up of two distinct elements that are not in direct contact with each other, ie., there is a gap between them with no structure to represent the bottom part of the V shape.
In addition, the transfer contact in Schwager is a rigid sphere that does not have the shape or flexibility of the Casler endless loop transfer contact. In Schwager, the spherical transfer contact is not, and cannot be, distorted into a kidney shape, does not make an appreciable arc of contact with the conducting elements, and is not wedged in place between the contact members by its own resiliency. Schwager’s rigid spherical transfer contact is kept in place, inter alia, by the adjustable, free-floating outer assembly referred to above, and by ball retainers (labeled element 106, located above and below the spherical transfer contact in Figure 9 above). Schwager does not suggest that his design wipes away impurities on the contact surfaces. If wiping inherently does occur in the Schwager device, it would result from very different force factors than those in Casler, where wiping is produced, at least in part, by the change in spacial orientation of the flexible loop as it enters and exits the grooves. Furthermore, Casler is able to produce a plurality of parallel paths of electrical contact with a single transfer contact because each coil of the Casler endless loop transfer contact maintains contact with the grooves. Schwager, on the other hand, would require a multitude of ball bearings to produce a plurality of contacts.
The Mathews Patent
The Mathews Patent, issued on April 9, 1957, and entitled “Bus-Bar,” describes a stationary electrical switch that is used to transmit current of low voltage and low amperage to a work circuit. The switch employs a flexible, helically wound'spring as a transfer contact. When the switch is turned on, the spring makes contact with a cylindrical bus bar, which has a plurality of V-shaped grooves machined into it. Mathews teaches that the combination of the spring and the plurality of V-shaped grooves results in multiple effective points of contact between the contactor and the bus bar.
Mathews also discloses that in operation, portions of the wire spread slightly and rub against the angular faces of the V-shaped grooves, and thereby increase the surface contact and keep the coil and grooves clean and polished. Mathews warns, however, against making the diameter of the coil so wide as to result in the coil being wedged in the V-shaped groove. Mathews indicates that the coil should have “perfect freedom of angular motion in both its circuit making and circuit breaking movements.”
Unlike the Casler claims in issue, Mathews is not concerned with the design of a rotary electrical connector, but rather with a stationary electrical switch. Hence, the conductive elements do not rotate. Moreover, in the switch, the transfer contact is not compressed between opposing V-shaped grooves in an inner and outer conducting member. Rather, the spring only touches one member at a time; contact is established by alternatively bending the spring element about one of two bus-bars. The Mathews spring is not distorted into a generally kidney-shaped form and, since the bus bar is straight and not round, the spring does not make a material arc of contact with the conductive elements.
The Windred Text
In his 1940 textbook entitled “Electrical Contacts,” G. Windred discusses what he *726refers to as finger and wedge contacts, which have extensive application in medium power circuit breakers. In a finger and wedge contact, a moving wedge-shaped member engages between two or more stationary contact fingers. The stationary fingers are arranged in pairs on either side of the moving contact. Springs are used to force the fingers and the wedge-shaped contact to engage, with one spring being used for each pair of fingers. The fingers are allowed sufficient movement to ensure constant contact between the fingers and wedge-shaped member. Windred teaches that such wedge contacts can create high individual contact pressures which keep resistance low. Windred does not teach the use of grooves in the contact surfaces and does not employ an endless loop intermediate transfer contact.
Secondary Considerations
Plaintiff rests his contention that secondary considerations demonstrate the nonob-viousness of his invention almost entirely on the long and expensive effort by Sperry, described infra, to design a successful rotary electrical connector. There are no other significant instances of alleged infringement or commercial success and no other secondary factors which would tend to indicate that Casler’s invention was not obvious. See, e.g., Environmental Designs, 713 F.2d at 697-98, 218 USPQ at 869. Plaintiff has never received any financial returns from the '727 patent.
C. Analysis and Conclusion — The Claims in Issue are not Obvious in View of the Prior Art
Defendant has failed in its effort to overcome the statutory presumption of validity because it has not proved that the subject matter “as a whole would have been obvious at the time the invention was made to one of ordinary skill in the [relevant] art.” 35 U.S.C. § 103.
Lindenblad was concerned with high-voltage rather than low-voltage transmission. The essence of defendant’s invalidity argument is, in effect, that if a person having ordinary skill in the relevant art were designing a rotary electrical connector for low-voltage transmission, it would be obvious in view of the cited prior art to modify the Lindenblad design in such a manner as to produce the rotary connector design covered in Claims 1, 3, and 4 of the ’727 patent. But absent the benefit of hindsight, such a modification simply would not be obvious.
Aside from Lindenblad, none of the prior art relied upon by defendant involves rolling ring technology and, hence, none specifically addresses the types of problems confronted in rolling ring connector design. The other three references arguably teach the use of V-shaped grooves in some part of an electrical switch or an electrical connector, but that alone clearly is not enough. “There must be something in the prior art as a whole to suggest the desirability, and thus the obviousness, of making the combination” of the Lindenblad design and the V-shaped grooves of the other references. Interconnect Planning Corp. v. Feil, 774 F.2d 1132, 1143, 227 USPQ 543, 551 (Fed. Cir.1985) (quoting Lindemann Maschinenfabrik GMBH v. American Hoist & Derrick Co., 730 F.2d 1452, 1462, 221 USPQ 481, 488 (Fed.Cir.1984)). Herein, the prior art does not suggest modifying the Linden-blad design to employ V-shaped grooves or that such a modification could achieve the advantages of the Casler connector for low-voltage electrical transmission. See Interconnect, 774 F.2d at 1143, 227 USPQ at 551.
Schwager is the only one of the four references relied upon by defendant that arguably teaches a rotary connector that employs a transfer contact placed between V-shaped grooves. But as explained above, unlike Casler, the grooves are not in the conductive elements and, in any event, the transfer contact is a rigid sphere rather than a flexible loop and operates in a very different manner than Casler’s contact.4 Even under Jacobson’s interpretation, *727Schwager simply would not suggest to one of ordinary skill in the relevant art the benefits that would result from forming V-shaped grooves in the conductive elements of Lindenblad and wedging a flexible transfer contact between such grooves.
Mathews, like Casler and Lindenblad, uses a helically wound connector. Mathews also discloses that his device has the benefit of self cleaning. But, as summarized above, Mathews was not concerned with the design of a rotary connector, did not involve mutually rotatable elements, did not use opposing V-shaped grooves, and did not compress the helically wound con-tactor between inner and outer conductive elements. Moreover, rather than suggesting any benefit from wedging, Mathews warned against it. In addition, the wiping that occurred in Mathews resulted from a very different structure and method of operation than that employed in Casler. Mathews simply would not suggest to one of ordinary skill in the relevant art the desirability of modifying the Lindenblad design so as to produce the rotary connector covered by the Casler claims.
The Windred text, which employs neither opposing V-shaped grooves nor a flexible loop contact, adds little of significance to the analysis of the obviousness of the Cas-ler claims.
In view of the above, the claimed invention would not be obvious when the teachings of Lindenblad are evaluated along with the teachings of Schwager, Mathews, and Windred. The prior art, when considered together and as a whole, does not suggest the benefits that resulted from Casler’s design or otherwise render obvious the structures of the rotary connector covered in Claims 1, 3, and 4 of the ’727 patent.5
*728Having determined that plaintiffs patent is not invalid, the focus next switches to the issue of infringement. The development and structure of the alleged infringing Sperry rotary electrical connector will first be described. Thereafter, an infringement analysis will be performed to determine whether the accused device infringes the ’727 patent.
IV. The Alleged Infringing Device
A. Sperry’s Initial Efforts Leading to its Own Patent Covering A Rolling Ring Rotary Connector
In February 1972, Sperry aerospace engineers began to experiment with varying designs of rotary electrical connectors in an effort to eliminate, or at least greatly reduce, the torque that was created when slip ring technology was employed. In December 1973, two Sperry engineers, Harold L. Swartz and William Stimson, filed a patent application covering various rolling loop electrical connectors. The two scientists were unaware of the related work described in the Lindenblad and Casler patents, and were under the impression that they had pioneered the rolling loop electrical connector.
The Swartz/Stimson patent application disclosed a variety of groove shapes and loop distortions. The loop distortions varied from virtually circular to a highly distorted kidney shape and the groove configurations included box, ripple, V-shaped, convex, and shallow, concave grooves. Pri- or to filing the application, Swartz and Stimson had tested only the use of a flat groove; the other groove shapes depicted in the application, including the V-shaped groove, were the product of a brainstorming session between Swartz and Stimson. The patent application indicated that rolling loop connectors as a class have far less sliding friction and torque than slip rings, but there was no detailed discussion of the relative merits of each of the disclosed groove configurations.
The patent examiner rejected the claims of the Swartz/Stimson patent application for failing to satisfy the nonobviousness requirement of 35 U.S.C. § 103. The examiner based the rejection on prior art that included the ’727 patent. After receipt of the Patent Office rejection, the Swartz/Stimson patent application was abandoned.
Sperry engineers, however, continued developmental work on rolling loop technology in an effort to develop a low-torque, low-friction rotary connector that could be used in sensitive aircraft and space instruments, including gyroscopes. Between 1973 and 1977, Peter Jacobson, a Sperry engineer, tested a variety of loop materials, loop deformations, and flexure preloads and also developed mathematical models for the design of rolling loop electrical connectors.
This work resulted in a January 14,1977, patent application by three Sperry engineers — Swartz, Jacobson, and Robert L. Pirman. The application covers a rolling ring electrical connector for use in highly sensitive instruments, such as for transmission of electricity between the relatively rotatable gimbals of gyroscope instruments. The specification discloses that previous efforts using slip rings as rotary connectors in gyroscopes had been plagued by manufacture and service problems, including noise, high friction-induced torque, and the buildup of friction polymers.6 The specification then describes a roll ring electrical connector that allegedly solves these problems in that the connector is characterized by substantially zero friction and coupling torque, no friction polymer buildup, relatively consistent current conduction, even in a shock and vibratory environment, and long life and low manufacture and assembly costs.
The roll ring connector described in the specification has the same three basic elements as the Casler and Lindenblad rotary connectors — inner and outer conductive elements positioned concentric and coplanar *729with a flexible loop placed between the rings as a transfer contact. However, the groove profile in the conductive elements is described as having a relatively shallow, arcuately concave configuration rather than Casler’s V-shape. In addition, rather than employing Casler’s stationary contact between the transfer contact and the grooves of the conductive elements, the device apparently permits temporary misalignment and self-capture as the transfer contact rolls between the conducting elements. The specification indicates that the relationship between the diameter of the transfer contact and the gap between the rings must be such to result in substantially no friction during the operation of the connector. The specification distinguishes the connector described in the Sperry patent application from the connector described in Casler’s ’727 patent, as follows:
It is quite evident that the device disclosed in [the Casler patent] is entirely unsuitable for use in applications which require substantially zero friction and coupling torques to be imposed on the supported rotatable member by the current transfer assembly because of the substantial wiping or rubbing friction and uncompensated bending moments generated as the coil or band enters and leaves the “V” grooves.
The Patent Office initially rejected the Sperry patent application on the ground that the claimed structure was obvious in view of the structures disclosed in the ’727 patent and one other patent. The patent applicants contested this rejection and, ultimately, the Patent Office agreed with them. On July 4, 1978, the three Sperry engineers were granted United States Patent No. 4,098,546 covering their roll ring connector design.7
B. Commercial Development of the Sperry Roll Ring
' In September 1978, after failure of the slip rings on NASA’s Seaset A satellite, scientists at NASA’s Jet Propulsion Laboratory became concerned about the electrical connectors in the Galileo Spacecraft, which was being constructed to probe the atmosphere of Jupiter. The scientists questioned whether roll rings should be used instead of slip rings to connect the rotating antenna to the stationary housing in the Galileo Spacecraft’s spin bearing assembly. However, since the vacuum environment in space greatly increases friction in any moving electrical connection, the scientists were concerned about fatigue failure, particulate debris, and cold welding of moving parts if a roll ring connector were used in the antenna connection. Therefore, the scientists established stringent requirements for the spin bearing assembly connectors, which covered fatigue, friction, linearity, noise, and frequency response.8
Sperry was awarded NASA Research and Development Contract No. 955,431 to build and test a spin bearing assembly and a standard articulation system for the Galileo Spacecraft. An initial decision was made to try to use roll rings instead of slip rings to connect the elements of the spin bearing assembly. Since available roll rings could not meet the rigorous specification requirements, it was recognized that development and evaluation of the roll ring concept would be required.
Sperry scientists performed extensive mechanical and electrical tests of roll ring systems.9 The only performance specifica*730tions that were not met from the outset related to the noise level — intermittent noise spikes — and to the life span of the connector. The life span problem, which had resulted from flexure fatigue, was ultimately solved, but the noise spike problem proved to be more difficult. Some scientists postulated that the noise spikes were caused by friction resulting from the roll ring transfer contact walking up the groove and then sliding back down. Although it was originally thought that the connectors must be absolutely free from organic contaminants or nonconductive elements of any kind, a lubricant was applied to reduce rolling friction and discourage any walking. The noise spike problem appeared to be solved by the use of a lubricated gold roll ring transfer contact, but there was insufficient time to test the system before the scheduled launch of the Galileo Spacecraft. As a result, in December 1980, slip rings were substituted for the proposed roll rings in Galileo’s spin bearing assembly.
No roll ring systems were ever delivered to the government pursuant to NASA Contract No. 955,481. However, the roll ring connectors developed by Sperry for the Galileo project were used in certain other projects by the United States which are classified. Evaluation of the extent of these other uses and related accounting issues have been deferred pending resolution of the issues of validity and coverage of the ’727 patent.
C. Structure of the Accused Device
The following diagram depicts the configuration of the Sperry Roll Ring Assembly which was used by the United States in various classified projects and which is alleged to be covered by plaintiff’s ’727 patent.
[[Image here]]
In the accused device, the grooves in the rotor and stator are shallow and concave in shape. The flexure is a flat-band loop that is extremely thin, i.e., with a thickness measured in thousandths of an inch. When the flexure is placed between the conductive elements, it is only slightly preloaded and still assumes a nearly perfect circular *731shape. The edges of the flexure are radi-used to nearly conform to the radius of the shallow, concave grooves. The following diagram depicts the configuration of the concave grooves on each of the conductive elements.
[[Image here]]
In operation, the Sperry connector permits temporary misalignment and self-capture of the transfer contact, as does the connector described in the Swartz-Jacobson-Pir-man patent.
V. The Infringement Analysis
28 U.S.C. § 1498 authorizes a patent owner to recover just compensation for unauthorized use or manufacture by the United States of inventions “covered by a patent,” i.e., covered by the patent claims. The legal and factual analysis employed to determine such unauthorized use or manufacture under Section 1498 parallels the two-step analysis used to determine “infringement” under 35 U.S.C. § 271(a). Lemelson v. United States, 752 F.2d 1538, 1548, 224 USPQ 526, 531 (Fed.Cir.1985). See, e.g., Teledyne McCormick Selph v. United States, 214 Ct.Cl. 672, 682-85, 558 F.2d 1000, 1005-07, 195 USPQ 261, 265-67 (1977); Palumbo v. Don-Joy Co., 762 F.2d 969, 974, 226 USPQ 5, 7-8 (Fed.Cir.1985). First, the scope of the relevant claims is ascertained, and second, a determination is made as to whether the claims as so interpreted cover the accused device. Id.
A. Scope of the Claims — Legal Standards
The patent specification contains “a written description of the invention, and of the manner and process of making and using it.” The specification concludes with one or more claims that “particularly point[ ] out and distinctly claim[ ] the invention.” See 35 U.S.C. § 112. When drafting the specification and claims, the patentee has a “verbal license”; the patentee is, in effect his or her own lexicographer and is free to define the claim terms in any way he or she chooses. W.L. Gore & Assocs., Inc. v. Garlock, Inc., 721 F.2d 1540, 1558, 220 USPQ 303, 316-17 (Fed.Cir.1983), cert. denied, 469 U.S. 851, 105 S.Ct. 172, 83 L.Ed.2d 107 (1984); Fromson v. Advance Offset Plate, Inc., 720 F.2d 1565, 1569, 219 USPQ 1137, 1140 (Fed.Cir.1983) (quoting Autogiro Co. of Am. v. United States, 181 Ct.Cl. 55, 62, 384 F.2d 391, 397, 155 USPQ 697, 702 (1967)); Lemelson v. United States, 14 Cl.Ct. 318, 322, 6 USPQ2d 1657, 1660 (1988). The fundamental issue, therefore, when determining the scope of a disputed claim is how the inventor defined the terms in the claim. Id. Resolution of this issue is approached from the perspective of one skilled in the relevant technological area. Id.
In resolving this fundamental issue, the court is obliged to consider three sources of information: (1) the patent specification; (2) the other claims in the patent; and (3) the patent’s prosecution history. Howes v. Medical Components, Inc., 814 F.2d 638, 643, 2 USPQ2d 1271, 1273 (Fed. Cir.1987); Moeller v. Ionetics, Inc., 794 F.2d 653, 656, 229 USPQ 992, 994 (Fed.Cir.1986). The specification is “the primary basis for construing the claims,” Standard Oil Co. v. American Cyanamid Co., 774 F.2d 448, 452, 227 USPQ 293, 295-96 (Fed. Cir.1985), because 35 U.S.C. § 112 provides that the specification shall contain “a written description of the invention ... in full, clear, concise, and exact terms.” The other claims in the patent are potentially important because differences between those claims and the claim in issue can provide crucial clues to claim interpretation. For example, the doctrine of claim differentiation raises a rebuttable presumption that differences in claims are significant and, therefore, that narrow limitations contained in one claim should not be read into other claims in which there is no such limitation. See Tandon Corp. v. United States Int’l Trade Comm’n, 831 F.2d 1017, 1023-24, 4 USPQ2d 1283, 1288-89 (Fed. Cir.1987); D.M.I., Inc. v. Deere & Co., 755 F.2d 1570, 1574, 225 USPQ 236, 239 (Fed. Cir.1985). Consideration of the prosecution history of the patent is required because *732“the prosecution history ... limits the interpretation of claims so as to exclude any interpretation that may have been disclaimed or disavowed during prosecution in order to obtain claim allowance.” Standard Oil, 774 F.2d at 452, 227 USPQ at 296.
In addition to the specification, other claims, and prosecution history, when interpreting claim terms, the court may consider other forms of extrinsic evidence such as expert testimony, see, e.g., Howes, 814 F.2d at 643, 2 USPQ2d at 1273; Moeller, 794 F.2d at 656-57, 229 USPQ at 994-95; Lemelson, 14 Cl.Ct. at 326, 6 USPQ2d at 1663, or texts such as dictionaries which explain common usage of the disputed claim terms. See, e.g., Fromson, 720 F.2d at 1571, 219 USPQ at 1141.
Herein, the parties disagree only with respect to the proper interpretation of certain claim terms. These disputed terms were underlined in the text of the claims at pages 6-7, supra.
At trial, each party presented expert testimony to support its respective proposed definition of each disputed claim term. Plaintiff testified as his own expert, and defendant employed two Sperry scientists, Harold Swartz and Peter Jacobson, as its experts at trial. Both of defendant’s experts worked on the development of the alleged infringing Sperry connector. Defendant also submitted dictionary definitions from the Random House Dictionary of the English Language (1966) for certain of the disputed claim terms. In the following sections, for each claim term that is in dispute, relevant portions of the specification, other claims, prosecution history, and other extrinsic evidence submitted at trial will be described. Then the evidence will be analyzed and the disputed claim term will be defined.
B. Assessment of the Scope of the Claims — Definition of the Disputed Terms
1. “V-Shaped Grooves” (Claim 4) and “Generally V-Shaped Grooves” (Claims 1 and 3)
Plaintiff contends, in effect, that the terms “V-shaped groove” (Claim 4) and “generally V-shaped groove” (Claims 1 and 3), used to describe the shape of the grooves in the conductive elements, are identical in meaning. He contends that each term delineates a groove of “any shape designed and spaced to compressively engage and guide a rolling loop flexure so that rolling contact takes place at the sides and not the bottom or center portions of the grooves and mating loop flexures.” Thus, plaintiffs definition would focus exclusively on the seating arrangement of the transfer contact in the grooves of the conductive elements and would not require that the grooves otherwise have any particular appearance.
Defendant contends that the terms “V-shaped groove” and “generally V-shaped groove’’ differ in scope and describe the particular appearance of the grooves of the conductive elements. Defendant contends that a V-shaped groove is a steep, straight-’ sided circumferential channel shaped like the letter V with an acute angle small enough to impose wedging, and that a generally V-shaped groove would have the same limitations except that the sides could be somewhat contoured.
Specification
In the specification, the grooves of the conductive elements of the rotary connector are depicted in Figures 1, 8, and 9. Figure 1 depicts grooves with two straight sides meeting at a sharp point. Figures 8 and 9 depict grooves in which the apex is rounded. In Figure 8, the sides are essentially straight, and in Figure 9, which the specification refers to as a “contoured form of a V-groove,” the sides are somewhat contoured.
*733[[Image here]]
The specification describes the grooves in the Casler invention as follows:
[T]he sides of the grooves, in either or both of the contacts ... can be contoured, somewhat as shown through the section ... of a contact ring illustrated in Figure 9. Such contouring increases slightly the area of contact at the expense of contact pressure. The apex of the groove need not be carried to a sharp point. It may be rounded as shown in this same figure as long as the groove is carried deep enough and the radius at the apex is enough smaller than that of the helical turns that contact takes place at the sides and not at the bottom of the groove.
In addition, the specification teaches that the Y-shaped or generally Y-shaped structure of the grooves is crucial to the performance of the Casler invention. As described below in the discussion of the disputed claim term “wedged,” wedging a transfer contact between V-shaped grooves results in a mechanical advantage that produces high-contact pressure between the transfer contact and the grooves of the conductive elements. The specification indicates, in effect, that the V-shape of the grooves is integral to securing this high-contact pressure.
With respect to defendant’s contentions that the angle at the apex of the groove must be acute, the angle depicted in Figures 1, 8, and 9 are acute (i.e., less than 90 degrees). In addition, the specification makes three specific references to an acute angle at the apex of the groove. First, in describing one of the figures, the specification states the relationship between the angle at the apex of the groove and the wedging action, as follows:
The more acute the angle of the V the greater the mechanical advantage and the higher the contact pressure, so that even though the radial pressure may be relatively light the contacting pressure is several times as great.
Second, in describing an embodiment of the Casler stationary switch, and comparing that design to the design of the Casler rotary connector, the specification states:
In the illustrative embodiment [of the stationary switch] disclosed, the parallel contacts are a pair of flat strips disposed in planes that intersect at an acute angle outwardly from the location of the transfer contact; the V-shaped groove so formed is quite similar to those formed in the rotor and stator of the other embodiment.
(Emphasis added.)
Third, in describing the rotary connector depicted in Figures 2-5, the specification states:
If the angle at the apex of the groove is acute and the sides correspondingly steep, so that the mechanical advantage is great and the contact pressure high, the well-known ‘irreversible’ property of a wedge comes into play and the turns remain firmly seated throughout the arcs around the inner and outer grooves where they are relatively at rest, the contact pressure being exerted by the turns themselves in their efforts to resume their truly circular form.
Other Claims
Claims 6-10, which cover the Casler stationary switch rather than the Casler rotary connector, employ the term “generally V-shaped groove” and, in addition, specify that the angle of the groove is “an acute angle.” Claims 1, 3, and 4, which cover the Casler rotary connector, use the terms “V-shaped” (Claim 4) or “generally V-shaped” (Claims 1 and 3) to describe the grooves *734but do not contain the additional limitation that the angle at the apex of each groove is acute. Thus, pursuant to the doctrine of claim differentiation, there is a presumption that the “acute angle” limitation included in Claims 6-10 should not be read into Claims 1, 3, and 4. See Tandon, 831 F.2d at 1023-24, 4 USPQ2d at 1288-89; D.M.I., 755 F.2d at 1574, 225 USPQ at 239.
Prosecution History
During the course of the patent prosecution, the Patent Office rejected Claims 6-10 as unpatentable over the Linn and Linden-blad patents. The Linn patent, described above in n. 5, employs a spring to conduct electricity between two contact members with elliptical-shaped grooves. In rejecting Claims 6-10, the Patent Office contended that, like the connector described in Claims 6-10, Linn “also shows generally V-shaped grooves in the contact members.”
In his response, plaintiff disagreed with the Patent Office’s characterization of the Linn grooves and stated that Linn did not show contact members having sections that form a generally V-shaped groove. The apparent rationale for this statement is contained in a contemporaneous letter to plaintiff from his patent counsel. In that letter, counsel expresses optimism that the Patent Office's rejection of the claims could be overcome because “Linn’s grooves are curved and not of a ‘V’ nature, which would cause the spring to wedge as disclosed in this present application.” Plaintiff prevailed with his response, and the Patent Office allowed Claims 6-10. The prosecution history thereby indicates that plaintiff did not intend the term “generally V-shaped groove” to encompass elliptical grooves of the type disclosed in Linn.
Expert Testimony
Testifying as an expert, plaintiff supported his position that the claim terms “V-shaped groove” and “generally V-shaped groove” cover any groove shape that results in side rather than bottom contact between the transfer contact and the grooves of the conductive elements. Plaintiff based his testimony primarily on the specification statements that the sides of the grooves can be contoured and the bottom rounded, and that the transfer contact makes side rather than bottom contact with the grooves. In addition, plaintiff stressed that Claims 1, 3, and 4, unlike Claims 6-10, do not contain any limitation requiring that the angle at the apex of the groove be an acute angle.
Defendant’s expert witness, Jacobson, interpreted the relevant evidence to support defendant’s contention that the term “V-shaped groove” covers a steep, straight-sided circumferential channel shaped like the letter V with an acute angle small enough to impose wedging, and that the term “generally V-shaped groove” covers the same structure with somewhat contoured sides. Jacobson relied primarily on the specification descriptions of the grooves and on plaintiff’s statement during the course of the patent prosecution history to the effect that Linn’s elliptical grooves were not generally V shaped.
Analysis
Neither plaintiff’s nor defendant’s proposed definition is adopted. Plaintiff’s proposed definition is deficient in two basic respects. First, “[tjhere is presumed to be a difference in meaning and scope when different words or phrases are used in separate claims.” Tandon, 831 F.2d at 1023, 4 USPQ2d at 1288. Plaintiff’s proposal to use a single definition for both V-shaped groove in Claim 4 and generally V-shaped groove in Claims 1 and 3 ignores the difference in the claims, i.e., ignores the use of the adjective “generally” in Claims 1 and 3, but not in Claim 4.
Second, and more fundamentally, plaintiff errs in contending that the claim terms should be interpreted to encompass “any shape” groove that performs the function of producing straddled contact with the transfer contact. The claims do not call for “any shape” groove that performs this function but, to the contrary, require particular shapes — V-shaped and generally V-shaped, i.e., shaped like or shaped generally like the letter V — the 22nd letter in the English alphabet.
A groove can provide straddled contact with the roll ring without generally having the appearance of the letter V. For exam-*735pie, the relatively shallow, arcuate grooves disclosed in the Sperry patent provide straddled rolling contact with the roll ring but do not generally have the appearance of the letter V. Plaintiffs contention to the effect that such a groove can be V shaped even if it does not have the shape of the letter V is inconsistent with the specification. Consistent with the appearance of the letter V, the specification indicates, in effect, that while the generally V-shaped grooves can have some degree of contour, the grooves have sides which meet at an apex. The three figures in the specification that depict the grooves, while having somewhat varying shapes, all generally have the appearance of the letter V.10 In addition, plaintiffs statement to the Patent Office to the effect that the elliptical opposing grooves of the conductive elements in the Linn patent are not V shaped also supports the conclusion that plaintiff was concerned in part with the appearance of the groove when he used the terms “V shaped” and “generally V shaped.”
Turning next to defendant’s proposed definition, defendant is incorrect in its contention that the term “V-shaped groove” necessarily covers only grooves with straight edges. By referring to the groove in Figure 9 as a “contoured form of a V-groove,” the specification indicates that a groove can have a shape like the letter V without having absolutely straight edges. Moreover, this court takes judicial notice that the letter V can be written in a variety of ways, some of which, such as script, do not necessarily use straight sides.
A more complex issue of claim interpretation, however, is presented by defendant’s contention that the terms “V-shaped groove” and “generally V-shaped groove” require acute angles at the apex of the grooves. Certain factors militate against including such a limitation. First, a groove can have the appearance of the letter V even if it has an angle of somewhat greater than 90 degrees, e.g., . As a general rule of claim interpretation, limitations not specifically present in claims should be read into the claims only when required, i.e., only when the specification, prosecution history, or other extrinsic evidence demonstrates that the terms actually present in the claim should be interpreted to encompass a term not specifically present. See, e.g., E.I. Du Pont de Nemours & Co. v. Phillips Petroleum Co., 849 F.2d 1430, 1433, 7 USPQ2d 1129, 1131 (Fed. Cir.1988); Specialty Composites v. Cabot Corp., 845 F.2d 981, 987, 6 USPQ2d 1601, 1605 (Fed.Cir.1988).11 Second, as noted above, since the term “acute angle” is used in describing the V-shaped grooves in Claims 6-10, under the doctrine of claim differentiation, a presumption results that the term “acute angle” should not be read into Claims 1, 3, and 4.
In this context, unless the specification unambiguously requires interpreting Claims 1, 3, and 4 as limited to grooves with acute angles, such a limitation should not be read into the claims. The specification, however, is at least arguably ambiguous on this point.
The specification clearly indicates that the angle at the apex must be sufficiently small to produce the desired wedging action. However, the specific references to acute angles in the specification and the use of acute angles in each relevant figure *736in the specification, while strongly suggesting that wedging will result only if the angles are acute, are at least arguably ambiguous.12 In view of this arguable ambiguity, rather than requiring that the angle necessarily be acute to be V shaped, it is appropriate only to require that the angle be small enough to produce wedging, as that claim term is defined in the next section of this opinion.13
Hence, the evidence submitted at trial indicates that the claim term “V-shaped groove” should be defined as a groove that (1) has the appearance of the letter V with two sides meeting at an apex, and (2) has an angle sufficiently small to result in the transfer contact being wedged when located between the opposing grooves of the conductive elements. The term “generally V-shaped groove” is somewhat broader in scope in that it covers a groove that “generally” has such a V shape.
2. “Wedged” (Claims 1, 3, and k)
Plaintiff contends that a transfer contact is wedged between the grooves of the conductive elements whenever the contact is captured on its sides in the V-shaped grooves and the interaction between parts is free rolling rather than a binding restraint against rolling. Thus, in effect, plaintiff alleges that the only pressure that must exist between the transfer contact and the grooves of the conductive elements for the transfer contact to be wedged is a pressure sufficient to capture the transfer contact on the sides of the V-shaped grooves. Defendant contends, in effect, that this minimal pressure is not sufficient and defines the claim term “wedged” to mean “to thrust, drive or pack tightly.”
Specificiation
The specification attributes much of the effectiveness of the Casler device to the wedging of the transfer contact in the grooves of the conductive elements. Wedging is a result of the interaction between the V-shaped or generally V-shaped grooves and the transfer contact. Wedging is produced by the natural tendency of the transfer contact, when distorted into a kidney shape, to seek to return to its original circular shape while in place in the grooves. In this regard, the specification states:
The inherent resilience of this spring contact wedges it into the V-shaped grooves so that the unit pressure between the sides of the grooves and the lines of contact formed between the sides of the grooves and the spring can be greatly in excess of the diametric force exerted by the spring in its tendency to assume its natural form as a circular toroid.
(Emphasis added.)
By producing high-unit contact between the transfer contact and the grooves, wedging assures that resistance is low and that the flow of electricity is uninterrupted. High-unit pressure results both in the grooves being wiped clean by “scraping of the surfaces where the spring enters and leaves the groove,” and in the spring being “stationary relative to the grooves when in operation,” i.e., there is “no slippage,” and the spring cannot be displaced by vibration.
The force resulting from wedging is of such magnitude that when a coil spring is employed as the transfer contact, the coils are flattened to some extent when in the V grooves. The specification indicates that *737the transfer contact is “snapped” into place, and that:
As the turns of the spring wedge into the grooves at the advancing of the loop their circular form is itself slightly distorted so that they make a short line contact with each side of the groove instead of the theoretical point contact, and this also serves to decrease the contact resistance.
As noted above, the specification also refers to the irreversible property of a wedge. The specification states:
If the angle of the apex of the groove is acute and the side correspondingly steep, so that the mechanical advantage is great and the contact pressure high, the well-known ‘irreversible’ property of a wedge comes into play and the turns remain firmly seated throughout the arcs around the inner and outer grooves where they are relatively at rest, the contact pressure being exerted by the turns themselves in their efforts to resume their truly circular form. There is therefore no measurable change in the contact resistance throughout the portion of their paths where they are relatively stationary with respect to the inner and outer grooves.
Finally, the specification refers to the extended lengths of grooves where the transfer contacts rest as “wedge seating.”
Other Claims
There is nothing in the other claims that is significant in defining the term “wedged.”
Prosecution History
Plaintiff made a number of representations during the course of the prosecution of the patent application that are important in determining how plaintiff defined the claim term “wedged.” First, plaintiff, in effect, quantified to some degree what he meant by the specification statement that the force resulting from wedging is “greatly in excess of the diametric force exerted by the spring in its tendency to assume its natural form as a circular toroid.” (Emphasis added.) During the prosecution, the Patent Office rejected Claims 1-5 as unpat-entable over the Lindenblad patent in view of the Boeye patent. The rejection stated that “[n]o invention is seen in substituting Boeye’s spring and groove type of transmission for Lindenblad’s flat spring in the event wiping contact is desired.” In his response to that rejection, plaintiff distinguished Lindenblad by stressing the high-unit pressure that wedging provides in the Casler invention. Plaintiff stated:
[Applicant’s structure provides multiple points of stationary, high unit pressure contact to achieve solid electrical connections. This high unit pressure is achieved because of wedging action of the helical spring into the Y-shaped grooves, so that the contact pressure is greater by several times than the spring pressure. Lindenblad’s structure does not permit the attainment of these high contact pressures, which make applicant’s structure advantageous and de-sireable for the accurate transmission of low energy electrical signals....
(First emphasis in original.)
Plaintiff also stressed the high-unit pressure resulting from wedging when he responded to the patent examiner’s citation to the Boeye patent and to a patent issued to Fjellstedt,14 both of which show a spring making side rather than bottom contact with the grooves. Plaintiff responded that neither reference was significant because neither achieved the high-unit pressure of Casler’s wedged device. These statements indicate that when plaintiff used the term “wedged” in the claim, he meant not only that contact occurs at the sides rather than the bottom of the grooves, but also that the contact was characterized by high-unit pressure.
Dictionary Definition
The relevant portion of the Random House Dictionary, presented into evidence by defendant, defines wedge, in pertinent part, as follows:
1. a piece of hard material with two principal faces meeting in a sharply acute angle, for raising, holding, or split*738ting objects by applying a pounding or driving force, as from a hammer. ... 2. a piece of anything of like shape: a wedge of pie. ... 4. ... an elongated area of relatively high pressure. 5. something that serves to part, split, divide, etc.10. to pack or fix tightly by driving in a wedge or wedges. 11. to thrust, drive, fix, etc., like a wedge_ ... 14. to force a way like a wedge_
Expert Testimony
To support defendant’s contention that the term wedged means “to thrust, drive or pact tightly,” defendant’s expert witness, Jacobson, relied primarily on the description in the specification and prosecution history of the high-unit pressure that results from wedging, and on the reference in the specification to the irreversible property of a wedge.
To support his contention that wedging requires only that the transfer contact is captured on the sides of the groove and the interaction between the parts is free rolling, plaintiff focused primarily on the specification’s description of the transfer contact in the grooves as “wedge seating.” He testified that wedge seating does not involve a locked state such as when an ax is wedged into wood, but rather merely “side seating.” He suggested, for example, that a wedge of cake is in side contact with other pieces of the cake. With respect to the specification reference to the irreversible property of a wedge, plaintiff indicated that it was either a typographical error or was improperly used. He described the irreversible property of a wedge as a locking action that would involve forcing the flexure into the groove to such an extent that force would be required to pull it out. Such locking action, he explained, would render a rotary connector inoperable.
Analysis
Again, neither definition offered by the parties is adopted in its entirety. Plaintiffs proposed definition of wedge is too broad in that it does not contain any limitation as to the amount of pressure exerted by the transfer contact on the grooves. As detailed above, the specification and prosecution history indicate that wedging results in high-unit contact pressure between the transfer contact and the grooves. The pressure must be “many times the spring pressure” inherent in the transfer contact seeking to return to its undistorted form, and the pressure must be sufficiently high to assure that when the connector is in operation, there is no slippage and the transfer contact remains stationary relative to the grooves and cannot be displaced by vibration.
Plaintiff’s analogy to the seating of a “wedge of cake” is simply misplaced. A wedge of cake is apparently referred to as a wedge because of its three-dimensional triangular shape (see Random House Dictionary, definition 2, supra), rather than because of its position merely touching adjoining pieces of cake.
Defendant’s proposed definition of wedged (“to thrust, drive or pack tightly”) is better than plaintiffs but must be more precise. The term “tightly” is a relative term and somewhat indefinite; what is tight to some may be loose to others. Since a determination will have to be made as to whether the transfer contact in the accused Sperry roll ring connector is wedged, it is appropriate, if possible, to quantify the term “tightly” in defendant’s definition. The specification permits such quantification through its description of the results of wedging, described above. In the context of these descriptions, a transfer contact should be deemed to be wedged between the grooves of the conductive elements when it is packed sufficiently tightly so that (1) the contact pressure between the transfer contact and the grooves is several times greater than the pressure inherent in the transfer contact attempting to return to its undistorted form, and (2) the transfer contact remains stationary relative to the grooves when the conductive elements are rotating.
3. Distorted into a “Generally Kidney-Shaped Form”
The parties’ respective proposed definitions of the term “generally kidney-shaped form” differ primarily with respect to the *739extent to which the transfer contact must be indented by the inner conductive element when compressed between the grooves of the conductive elements before it can be considered generally kidney shaped. The parties have stipulated that whenever the diameter of the transfer contact exceeds the radial separation between the grooves of the conductive elements, the convex outer surface of the inner conductive element will at least “tend to” indent a reverse curvature on the surface of the flexure. Plaintiff contends in effect, that such a tendency to indent is sufficient to constitute generally kidney shaped, defining that term to mean the “distortion of a normally circular loop assembled between concentric contact grooves with a radial spacing less than the free diameter of the loop.” Defendant contends, however, that a tendency to indent is not sufficient and defines generally kidney shaped to mean “deformed into the shape of a long oval indented at one side to exhibit a substantial clearly visible arc of reverse curvature.”
Specification
The specification specifically describes a kidney-shaped loop in three places, two of which make reference to Figure 2, which is set forth above. The specification states:
[The] transfer contact ... is a spring of toroidal form whose diameter [prior to being inserted between the V-shaped grooves] is greater than the radial separation between the grooves in the rotor and stator contacts. ... Owing to its diameter, when it is sprung into place to engage the grooves in both ... contacts, it is distorted into a generally kidney-shaped form that contacts a material arc within the groove of both ... contacts. ... The excess in diameter of the undistorted spring over the separation between the ... contacts ... insures that a material arc of both ... contacts is in contact with the spring.
******
FIG. 2 is a plan view of the apparatus shown in FIG. 1, partly in section, illustrating more clearly the kidney-shaped contour taken by the spring transfer-contact when in place....
******
The diameter of the toroid ... must exceed the separation between the grooves ... and preferably it is much greater than their separation ... the upper limit being that when the toroid is distorted and snapped into place as shown in FIG. 2, the two ends of the loop formed by it do not come into contact, although they may nearly surround the inner contact_ In general it is desirable that a large number of turns of the toroid come into contact with each of the grooves, so that a fairly large diameter is indicated but too large a diameter will sometimes cause the loop to pull away from the central portion of its contact with the inner groove.... The preferred diameter is such that when in place the spring is distorted from its originally toroidal form into the somewhat kidney-shaped form shown in FIG. 2. Here and in the claims that follow, however, “kidney-shaped” is intended to be construed broadly to include cases where the spring nearly surrounds the inner contact.
In addition to the statements above that specifically refer to the term “kidney-shaped,” other parts of the specification are also relevant in assessing the minimum amount of reverse curvature necessary for the wedged transfer contact to be deemed generally kidney shaped. In each patent figure that depicts the transfer contact in its distorted form, the transfer contact shows a clearly visible reverse curvature in its area of contact with the inner conductive ring. In addition, the specification indicates that high-unit pressure and wiping, which are critical to the Casler invention, are achieved through wedging the transfer contact “along extended lengths of the groove” and “along extended portions of the contacts.” The specification also indicates, in effect, that the arc of contact is sufficiently large to result in “many parallel, low resistance paths” between the relatively rotating parts.
To support its definition of generally kidney shaped, plaintiff places special reliance *740on the following statement in the specification:
The relative radii of the inner and outer contacts ... are unimportant, the criteria for satisfactory operation of the invention being that the loop diameter of the transfer contact be greater than the separation of the grooves and that this separation be constant throughout the extent of the grooves.
Other Claims
Claims 1 and 4 describe in different ways the shape of the transfer contact when wedged between the grooves of the conductive elements. Both claims indicate that the wedged contact is “distorted into a generally kidney-shaped form.” Claim 1 contains the additional description that “there is an appreciable arc of contact between the transfer contact and the contact members.” The use of the additional limitation to describe the transfer contact in Claim 1 is at least potentially significant when defining generally kidney shaped because under the doctrine of claim differentiation discussed swpra, a presumption results that differences in the claim terms are significant. Hence, there exists a presumption that the term “generally kidney-shaped form” is different in scope from the term “appreciable arc of contact.”
Prosecution History
The prosecution history indicates that in fact plaintiff intended the terms “appreciable arc of contact” and “generally kidney-shaped form” to be different in scope. When the original Casler patent application was filed, the term “generally kidney-shaped form” appeared in Claim 4 but not in Claim 1. Both claims provided that the diameter of the undistorted transfer contact was greater than the radial separation between the grooves of the conductive elements, but they described in different ways the distortion that resulted when the transfer contact was wedged between the grooves of the conductive elements. Claim 4 provided that the transfer contact was distorted into a generally kidney-shaped form, while Claim 1 provided that there was an appreciable arc of contact between the transfer contact and each of the conductive elements.
After allowing Claim 4, the Patent Office rejected Claim 1 on the ground that it covered a structure in which the transfer contact surrounded the inner contact member — a structure that would be inoperable. In response, plaintiff amended Claim 1 to include the term “generally kidney-shaped form” which was already contained in previously allowed Claim 4. The Patent Office then allowed the claim.
Thus, the prosecution history indicates that the term “generally kidney-shaped form” was added to Claim 1 to limit further its scope. It was intended to be different in scope than the term “appreciable arc of contact” in that it did not encompass a structure in which the transfer contact completely surrounded the inner contact member.
Dictionary Definition
The Random House Dictionary defines kidney shaped, in relevant part as “having the general shape of a long oval indented at one side.”
Expert Testimony
To support his contention that the wedged transfer contact could be generally kidney shaped even if it did not exhibit any visible indent, plaintiff focused on the specification statements that the diameter of the undistorted transfer contact “is greater than” and “must exceed” the separation between the grooves. Plaintiff contended, in effect, that so long as that requirement is satisfied, the resulting form of the transfer contact necessarily must be deemed to be generally kidney shaped.
In support of defendant’s contrary contention, Jacobson relied primarily on the patent figures which depict a clearly visible indent or arc of reverse curvature and on the above-cited portions of the specification that either use the term kidney shaped or otherwise describe the form that the transfer contact takes when it is wedged into place between the grooves of the conductive elements.
*741 Analysis
Defendant’s proposed definition of the term “kidney-shaped form” is essentially identical to the dictionary definition. As noted above, the Random House Dictionary defines kidney shaped as “having the general shape of a long oval indented at one side.” Defendant’s proposed definition, in essence, adds the qualification that the long oval indented at one side exhibits “a substantial clearly visible arc of reverse curvature.” But that additional qualification is inherent in the dictionary definition because a long oval with an indent at one side would necessarily display such a visible arc of reverse curvature.
A general rule of claim interpretation is that “[w]ords in a claim ‘will be given their ordinary and accustomed meaning, unless it appears that the inventor used them differently.’ ” Envirotech Corp. v. Al George, Inc., 730 F.2d 753, 759, 221 USPQ 473, 477 (Fed.Cir.1984) (quoting Universal Oil Products Co. v. Globe Oil & Refining Co., 137 F.2d 3, 6, 58 USPQ 504, 508 (7th Cir.1943), aff'd, 322 U.S. 471, 64 S.Ct. 1110, 88 L.Ed. 1399 (1944)). When a general descriptive term such as kidney shaped is used in a claim, the dictionary definition generally would constitute the ordinary and accustomed meaning of the term. See, e.g., Universal Oil Products Co. v. Global Oil & Refining Co., 40 F.Supp. 575, 582 (N.D.Ill.1941), aff'd, 137 F.2d 3, 58 USPQ 504 (7th Cir.1943), aff'd, 322 U.S. 471, 64 S.Ct. 1110, 88 L.Ed. 1399 (1944). Hence, the dictionary definition of kidney shaped should be employed unless it appears that plaintiff used the term differently. See, Envirotech, 730 F.2d at 759, 221 USPQ at 477. Herein, rather than suggesting a departure from the dictionary definition, the specification, other claims, and prosecution history each support its use. See, e.g., Fromson, 720 F.2d at 1569-71, 219 USPQ at 1140-41.
As described above, the specification indicates that the diameter of the undistorted transfer contact “is greater than” and “must exceed” the separation between the grooves, and that “the relative radii of the inner and outer contacts are unimportant, the criteria for successful operation of the invention being that the loop diameter of the contact be greater than the separation of the grooves.” But these statements do not specifically address the issue of how much greater the diameter of the transfer contact must be for the loop, when wedged, to be classified as generally kidney shaped. That issue is addressed in those portions of the specification, described above, which discuss the function or otherwise describe the arc of contact made by the transfer contact.
For example, the specification indicates that the high-unit contact pressure and wiping action that characterize the invention are achieved when the transfer contact engages the conductive elements “along extended lengths” or “extended portions” of the grooves. The arc of contact between the transfer contact and the grooves of the conductive elements is described elsewhere in the specification as “material,” and in other portions it is noted that the arc results in “many parallel low resistance paths” between the conductive elements. Moreover, the relevant figures in the specification that depict the transfer contact between the grooves each show large visible indentures in the sides of the transfer contact. Read in its entirety, it appears clear from the specification that plaintiff did not intend to stray from the ordinary and accustomed meaning of the term “generally kidney-shaped form.” Rather, plaintiff intended that the term define the general shape of a long oval with a substantial, visible indent in one side.
The other claims and prosecution history do not suggest to the contrary. As explained above, Claim 1 uses the term “appreciable arc of contact” in addition to the term “generally kidney-shaped” to describe the arc of contact, while Claim 4 uses only the term “generally kidney-shaped.” The prosecution history demonstrates that the term “generally kidney-shaped form” was added to Claim 1, which already contained the limitation “appreciable arc of contact,” only to limit the claim so that it did not read on a structure where the transfer contact completely surrounded the inner *742contact member. The dictionary definition of kidney shaped would not encompass such a structure and, hence, consistent with the doctrine of claim differentiation, would give different meaning to the terms “generally kidney-shaped” and “appreciable arc of contact.”
4. “Appreciable Arc of Contact”— Claims 1 and 3
As described above, in deriving a definition for “generally kidney-shaped,” the arc of contact between the transfer contact and the conductive elements is described in the specification in a number of ways. The term “appreciable arc,” however, is not used anywhere in the specification to describe the arc of contact. In selecting a claim term that does not appear in the specification, plaintiff presumably intended that it be interpreted in accordance with its dictionary definition.15 The Random House Dictionary defines the term “appreciable” to mean “capable of being readily perceived or estimated; considerable.” Using this definition, the claim term “appreciable arc of contact” would cover an arc that is capable of being readily perceived or estimated; an arc that is considerable. That definition is consistent with the description of the arc in the specification, other claims, and prosecution history, which is described above in connection with defining the term “kidney-shaped.”
5. A “Spring of Generally Toroidal Form” — Claim 1 and 3
The parties’ proposed definitions differ with respect to how thick the spring material forming the continuous loop must be for the loop to be considered of generally toroidal form. Plaintiff contends that spring material of any thickness whatsoever would be sufficient to produce a spring of generally toroidal form. Defendant contends that a spring of generally toroidal form must have a circular or radially thick cross-section.
Specification
The terms “toroid” and “toroidal” appear in the specification in connection with plaintiff’s description of its preferred embodiment of the transfer contact, which is the spring depicted as element 25 in Figures 2-5. Figure 5 shows the spring in undistorted form as follows:
[[Image here]]
The specification describes the spring in Figures 2-5 as follows:
Each of the contacts ... is in the form of a continuous loop or toroid of spring material. The preferred form, shown in FIG. 5, is a helically wound spring of highly resilient wire, such as beryllium-copper, with its ends brought together so that electrically and mechanically the spring forms a continuously toroidal helix.
The specification indicates that the spring need not necessarily have the helical loop shape shown in the preferred embodiment, but also could be a single loop of specified characteristics. The specification states:
It is also possible to use a toroidal spring formed of a single endless wire loop or a number of leaf springs. The springs need not be of circular cross-section; they may be elliptical or substantially rectangular in cross-section as indicated [in] ... FIG. 8_ If the loop is a solid wire ... it should be sufficiently thick, measured radially of the toroid, to be resistant enough to torsion so that, in the case of a loop of rectangular cross-*743section it will not tend to twist and make a flat contact against the sides of the groove, thus defeating the wedging action upon which much of the effectiveness of the device depends.
Other Claims and Prosecution History
As explained above, independent Claims 1 and 4 each cover a rotary connector made up of two conductive elements connected through the use of a spring transfer contact. Both claims describe the spring transfer contact using essentially the same series of limitations except that Claim 1 in addition calls for the spring to be of generally toroidal form. The prosecution history of the ’727 patent indicates that plaintiff intended this difference between Claims 1 and 4 to be significant. During the course of the prosecution, the patent examiner suggested that the patent title be changed by inserting the term toroidal so that the title read “Low-Resistance Toroidal Rotary Contact.” Plaintiff objected by indicating, in effect, that certain of the transfer contacts that could be employed in the invention would not be toroidal in shape. Plaintiff stated:
[T]he title of the present invention has not been amended to include the word toroidal, as suggested by the examiner, because it is believed that the definition of a toroid or toroidal cross section does not necessarily describe all the springs which might be successfully used in the combination low-current contact. For example, spring 25 in Figure 6 [Figure 8 of the patent] is not a toroidal shape and yet, as a part of the combination forming the electrical contact, it operates quite well.
The transfer contact depicted in Figure 8 appears to be a loop, consisting of a number of rectangularly shaped leaf springs.
Dictionary Definition
The Random House Dictionary defines toroid as “a surface generated by the revolution of any closed plane curve or contour about an axis lying in its plane.”
Expert Testimony
Plaintiff testified that the term toroidal means the form of the surface of a solid that is generated by revolving a closed surface about an axis remote from the closed surface, and hence covers loops of any thickness whatsoever. Jacobson, testifying for defendant, relied upon the portions of the specification, other claims, and prosecution history cited above to support defendant’s contention that to be classified as “of generally toroidal form,” a spring loop-must be of circular or radially thick cross-section.
Analysis
The term “toroidal” is descriptive of a particular shape. As noted above, the ordinary and accustomed meaning of a generally descriptive term is its dictionary definition. Hence, when such a term is used in a claim, the term should be defined according to its dictionary definition unless it appears that the patentee used the term differently. Herein, there is no indication in the specification, other claims, or prosecution history that a different definition was intended.
The dictionary definition of toroid, which is not materially different from plaintiff’s proposed definition, does not require that a loop have any minimum degree of thickness to be toroidal in form. The definition encompasses any “surface generated by the revolution of any closed plane curve or contour about an axis lying in its plane.” If a thin rectangular contour revolves about an axis lying on its plane, the toroid formed would be a thin continuous loop of rectangular cross-section.
Contrary to defendant’s contention, the specification and prosecution history do not reveal an intent on the part of the patentee to define toroidal differently. Defendant relies on the statement in the specification that “[i]f the loop is a solid wire ... it should be sufficiently thick, measured radially of the toroid ... so that, in the case of a rectangular cross-section, it will not tend to twist and make a flat contact against the sides of the groove.” But that description relates to the successful operation of the transfer contact in the disclosed Casler embodiments and does not suggest that a loop *744must be thick to fit within the definition of toroid, per se. Moreover, in stating that the wire should be “sufficiently thick, measured radially of the toroid ... to avoid twisting,” the specification indicates, in effect, that a loop could be toroidal in shape but not be “sufficiently thick” to avoid twisting.
Next, defendant purports to find support for its definition of toroid in plaintiff’s statement during the course of the prosecution that the leaf springs depicted in Figure 8 are “not a toroidal shape.” Defendant suggests that the leaf springs in Figure 8 are flat-band loops and that the flatness is the reason that plaintiff indicated that they were not toroid in shape.
But the figures and specification do not indicate that the leaf springs depicted in Figure 8 are necessarily flat, or that they are the equivalent of a flat-band loop. Moreover, there is no suggestion in the prosecution history that flatness was the basis for plaintiffs statement that the spring depicted in Figure 8 is not toroidal in shape. Plaintiff maintains that he objected to the use of toroidal in the title of his patent because springs other than those with toroidal shapes might be successfully employed in his invention. Plaintiff explained his illustrative reference to Figure 8 by stating that thin multiple structures, particularly those consisting of spiral windings, might not be considered typically toroidal in shape (as opposed to generally toroidal). Plaintiffs interpretation of toroidal is more easily reconciled with the specification than defendant’s in that the specification uses the descriptive term “toroidal” in reference to both helically wound spring loops and single loops of wire.
In this context, this court adopts the proffered dictionary definition as the proper construction of the term toroidal in the claim term “generally toroidal form.” Hence, a spring of generally toroidal form need not necessarily have a circular or radially thick cross-section to be toroidal.
C. Determination as to Whether the Claims Cover the Accused Device
The second step of the infringement analysis involves a determination of whether the claims as interpreted in the first part of the infringement analysis cover the alleged infringing device. Palumbo, 762 F.2d at 974, 226 USPQ at 7-8. A claim can cover a device in either of two ways. First, a device literally infringes a claim if the accused device embodies each and every element in the claim. Mannesmann Demag Corp. v. Engineered Metal Products Co., 793 F.2d 1279, 1282, 230 USPQ 45, 46 (Fed.Cir.1986); Builders Concrete, Inc. v. Bremerton Concrete Products Co., 757 F.2d 255, 257, 225 USPQ 240, 241 (Fed. Cir.1985) (citing Fay v. Cordesman, 109 U.S. 408, 420-21, 3 S.Ct. 236, 244, 27 L.Ed. 979 (1883)).16 Second, in the absence of literal infringement, infringement can be found under the doctrine of equivalents if the device “performs substantially the same overall function or works in substantially the same way, to obtain substantially the same overall result” as the claimed invention. Pennwalt Corp. v. Durand-Wayland, Inc., 833 F.2d 931, 934, 4 USPQ2d 1737, 1739 (Fed.Cir.1987), cert. denied, — U.S. -, 108 S.Ct. 1226, 99 L.Ed.2d 426, and cert. denied, Durand-Wayland, Inc. v. Pennwalt Corp., — U.S. -, 108 S.Ct. 1474, 99 L.Ed.2d 703 (1988). Under either literal infringement or infringement under the doctrine of equivalents, the patentee must demonstrate infringement by a preponderance of the evidence. See, e.g., Lemelson, 752 F.2d at 1547, 224 USPQ at 531.
Literal Infringement
Claim 1
Defendant contends that plaintiff has failed to establish literal infringement *745of Claim 1 because plaintiff has failed to demonstrate that the alleged infringing Sperry roll ring device contains a number of elements of Claim 1. Specifically, defendant contends that plaintiff has failed to satisfy his burden with respect to each of the claim elements, the definitions of which are in dispute. Defendant contends that plaintiff has failed to prove that the grooves of the conductive elements in the accused device are generally V shaped, and that the transfer contact in the accused device is “wedged” between the grooves of the conductive elements, is distorted into a “generally kidney-shaped” form, exhibits an “appreciable arc of contact,” and is of “generally toroidal form.” Based on an analysis of all of the evidence submitted at trial relating to the configuration as well as the operation of the accused device, the court concludes that plaintiff has not established literal infringement. The only one of these disputed elements plaintiff has proved is embodied in the accused device is a “spring of generally toroidal form.”
First, as defined above, to be generally V shaped, a groove must generally have the appearance of the letter V and in addition must have an angle sufficiently small to produce wedging. As explained above, the shallow, concave arc which serves as a groove in the Sperry device is an arc of a circle; it does not have two sides that intersect at an apex and cannot otherwise be said generally to have the appearance of the letter V.
Next, plaintiff has not demonstrated that the transfer contact in the Sperry device is wedged between the grooves of the conductive elements, as that term has been defined by the court. As explained above, to be wedged, the transfer contact must be packed sufficiently tightly that (1) the contact pressure between the transfer contact and the grooves is several times greater than the pressure inherent in the transfer contact attempting to return to its undistorted form, and (2) the transfer contact remains stationary relative to the grooves when the conductive elements are rotating. Plaintiff has not demonstrated that the accused device possesses either of these characteristics. In fact, the evidence indicates that it displays neither.
With respect to contact pressure, plaintiff acknowledged that the mechanical advantage in the shallow grooves of the Sperry connector used in the Galileo project provides only a 15-20 percent increase in contact force over the diametrical force of the springs, and that this limited level of mechanical advantage cannot be considered to be “greatly in excess” of the spring force. With respect to the position of the transfer contact in the grooves of the conductive elements during operation, the evidence indicates that rather than always remaining stationary relative to the grooves, the Sperry transfer contact can drift about in the grooves to some extent, with temporary misalignment and recapture.
Next, with respect to the degree of distortion of the transfer contact in the Sperry device, plaintiff has acknowledged that the loaded transfer contact in the Sperry device used in the Galileo project assumed a nearly perfect circular shape, and the depictions of the Sperry device in evidence support this same conclusion. A nearly perfect circular shape does not fit within the definitions derived above for either the claim term “generally kidney-shaped” (generally having the shape of a long oval with a substantial, visible indent at one side) or the claim term “appreciable arc of contact” (an arc that is capable of being readily perceived or estimated; an arc that is considerable).
Claim 3
Claim 3 is dependent on independent Claim 1 and, therefore, contains each of the limitations of Claim 1. The accused device therefore does not cover Claim 3 because it is missing the same elements as were found missing with respect to Claim 1.
Claim If.
Plaintiff has failed to establish literal infringement of Claim 4 for certain of the same reasons discussed above with respect to Claim 1. Claim 4, like Claim 1, requires that the transfer contact be wedged be*746tween the grooves of the conductive elements and distorted into a generally kidney-shaped form. As explained above, the Sperry device does not have either of these characteristics. In addition, Claim 4 is even narrower than Claim 1 to the extent that it requires that the grooves be “V shaped” rather than merely “generally V shaped.” As explained above, the grooves in the accused device are not generally V shaped, much less V shaped.
Hence, because plaintiff has failed to establish by a preponderance of the evidence that the accused device contains all of the elements in Claims 1, 3, or 4, plaintiff has not satisfied his burden of proof on the issue of literal infringement.
Infringement Under the Doctrine of Equivalents
The doctrine of equivalents is a judge-created doctrine devised to do equity “in situations where there is no literal infringement but liability is nevertheless appropriate to prevent what is in essence a pirating of the patentee’s invention.” Loctite Corp. v. Ultraseal Ltd., 781 F.2d 861, 870, 228 USPQ 90, 96 (Fed.Cir.1985). See also Texas Instruments, Inc. v. United States Int’l Trade Comm’n, 805 F.2d 1558, 1571-72, 231 USPQ 833, 841-42 (Fed.Cir.1986), reaff'd, 864 F.2d 1369, 6 USPQ2d 1886 (Fed.Cir.1988). As explained above, the doctrine permits a finding of infringement when the accused device “performs substantially the same overall function or works in substantially the same way, to obtain substantially the same overall result as the claimed invention.” Pennwalt, 833 F.2d at 934, 4 USPQ2d at 1739. The range of equivalents permitted a patentee under the doctrine of equivalents “depends upon and varies with the degree of invention.” Continental Paper Bag Co. v. Eastern Paper Bag Co., 210 U.S. 405, 415, 28 S.Ct. 748, 749-50, 52 L.Ed. 1122 (1908). A “pioneer” invention is entitled to a broad range of equivalents while an invention that makes less of a contribution is allowed a correspondingly smaller range of equivalents. See, e.g., Perkin-Elmer Corp. v. Westinghouse Elec. Corp., 822 F.2d 1528, 1532, 3 USPQ2d 1321, 1324 (Fed.Cir.1987); Triax Co. v. Hartman Metal Fabricators, Inc., 479 F.2d 951, 958, 178 USPQ 142, 147, cert. denied, 414 U.S. 1113, 94 S.Ct. 843, 38 L.Ed.2d 740 (1973).
Herein, plaintiff has apparently made a significant contribution to rotary connector technology, but the contribution is not properly characterized as “pioneer.” Lindenblad arguably made a pioneer contribution in the discovery of the rolling ring technology. Plaintiffs patented design, however, in essence constitutes a nonpio-neer modification of the Lindenblad design. In any event, however, even if plaintiffs invention was considered “pioneer,” resort to the doctrine of equivalents would not benefit him because the accused Sperry device does not perform substantially the same overall function or work in substantially the same way to achieve substantially the same overall result.
Viewed broadly, there are obvious and clear similarities in the work performed and the results obtained by the two devices. Both the patented invention and the alleged infringing device are rotary connectors, and both provide low-resistance electrical connection for transmitting low-amplitude signals. However, a closer analysis reveals substantial differences in the work performed and the results obtained. The Sperry connector was designed for particular functions in a space environment; a premium was placed on minimizing friction and coupling torques and on extending the life of the transfer contact by limiting pressures that could cause fatigue and failure of the transfer contact. The patented device, which employs high-pressure contact between a visually distorted transfer contact and V-shaped or generally V-shaped grooves, results in substantially greater friction and coupling torque and results in far greater pressures on the transfer contact. The patented connector would function poorly in the sensitive instruments for which the Sperry device was designed, and, in view of the comparatively little wiping involved, the Sperry device would function poorly in the contaminated environment for which the patented invention was apparently designed.
*747But even more apparent than the differences in function and result are the substantial differences in the way the two devices operate. The patented connector operates by wedging a transfer contact distorted into a generally kidney-shaped form between V-shaped or generally V-shaped grooves of the conductive elements. This combination results in high-unit contact pressure and an arc of stationary engagement between the transfer contact and the V-shaped grooves, and thereby prevents misalignment, produces a multiplicity of electrical paths along a kidney-shaped arc of contact, and provides for the wiping away of contaminants.
The Sperry connector, on the other hand, employs shallow, concave grooves, does not wedge the transfer contact between the grooves of the conductive elements, and does not produce a substantial visible reverse arc in the transfer contact. Rather than producing stationary contact, the loop in the Sperry device can roll freely subject to storing forces. The Sperry device maintains the transfer contact in a virtually circular rather than kidney-shaped form and thereby limits stress and extends the life of the transfer contact. Moreover, Sperry did not address the problem of the formation of nonconducting contaminant films by employing high-pressure wiping. Rather, Sperry applied gold plating and lubrication to the contacting loop and the groove surfaces and assembled the connectors in clean room facilities. If gold plating and lubrication were employed in Cas-ler’s rotary connector, they would be quickly penetrated and wiped away by the high-pressure, stressful interaction of the Casler distorted transfer contact and its generally V-shaped grooves.
In view of these differences in operation, the relevant elements of the Sperry device are not substitutable for the elements of the Casler device. For example, the shallow, concave grooves of the Sperry device would not produce the desired wedging if substituted for the V-shaped and generally V-shaped grooves in the patented device, and the transfer contact of the Sperry device, which remains virtually circular when loaded, would not produce the high-pressure and stationary contact if it remained virtually circular when loaded in the Casler device.
In sum, whether the comparison is made element by element or by viewing the devices as a whole, the two devices simply are not equivalent. In Perkin-Elmer, 822 F.2d at 1528, 3 USPQ2d at 1321, the Court of Appeals for the Federal Circuit warned that “a court may not, under the guise of applying the doctrine of equivalents, erase a plethora of meaningful structural and functional limitations of the claim on which the public is entitled to rely in avoiding infringement.” Id. at 1532, 3 USPQ2d at 1324. Plaintiff asks the court to do precisely that which is proscribed in Perkin-Elmer when it asks the court to find infringement herein under the doctrine of equivalents.
Conclusion
For the foregoing reasons, plaintiff has failed to establish that the ’727 patent covers the Sperry roll ring connector used by the United States. Judgment is therefore awarded to defendant. No costs.
IT IS SO ORDERED.
. The '727 patent was issued from Application Serial No. 317,110, filed on October 16, 1963, which was a continuation-in-part of Application Serial No. 98,443, now abandoned, filed on March 27, 1961, which itself was a continuation of plaintiffs original application, Serial No. 670,470, now abandoned, filed on July 8, 1957, entitled “Low-Resistance Rotary Contact."
. In its proposed findings of fact and post-trial brief, defendant relies exclusively on these four references.
. At the time of trial, Jacobson was an employee of Sperry.
. Schwager, like Lindenblad, apparently was primarily concerned with high-potential rather than low-power signal transmission.
. The same conclusion that defendant failed to prove invalidity results when the prior art relied upon by the patent examiner is also considered. Boeye is not at all concerned with an electrical connection; rather, it involves an assembly for use in transferring mechanical power from one shaft to a second shaft through the use of pulleys mounted on the shafts. The pulleys have identical steep-sided, flat-bottomed groove profiles located in their respective outer peripheries. Connecting the pulleys is a continuous helical spring having a cross-sectional profile in conformity with the groove profiles so as to engage the grooves along their sides and not their bottoms. In addition to not involving electric connection, the pulleys are not concentric and the grooves are not opposed throughout their entire circumferential length. The spring connector is positioned in tension, not in compression, and is not distorted into a generally kidney-shaped form. Unlike Casler, contact between the connector and the grooves is made over a relatively large area, e.g., the groove sides as opposed to edge contact as in Casler.
The Hopper apparatus is a stationary electrical switch which has a box-type groove in the outer periphery of the inner contact member. Helically wound coil is mounted on and completely surrounds the inner contact member and extends beyond the flanges of the groove. When the switch is activated by movement of a lever, the inner member and extended coil move axially, thereby permitting the coil to engage and thus establishing electrical contact with a stationary outer member. In Hopper, unlike in Casler, there are no moving elements other than the lever, there are no opposed V-shaped grooves, and the engaging elements are aligned coaxially. In addition, the spring is not com-pressively distorted into a kidney shape but rather stretched, in tension, about only one of the two contact elements.
Linn discloses a stationary, nonrotational electrical device for use in vapor-filled rectifiers of the sort used in mercury vapor lamps. The Linn apparatus has three elements. The first element is a housing in the general shape of a cylindrical canister. The canister is open at one end with an anode electrically insulated from the housing attached at the opposite end. The connecting surface on the open end of the housing has a groove on its inner periphery. The second element, a sealed chamber filled with metallic vapor, has a similar groove in its outer periphery with the same profile as the groove in the housing. Stretched about the chamber groove is a pair of resilient helical springs in the shape of half moons. The springs have a profile that conforms to that of the grooves, which establishes contact along the length of the coil. When assembled, the chamber is positioned inside the housing such that the grooves are opposed and the only contact between the chamber and the housing occurs where the spring elements engage both grooves. When in operation, the spring contactor provides electrical contact between the chamber and the housing. The opposed grooves may virtually abut one another or they may be separated by a gap. Thus, unlike Casler, the device in Linn does not have any moving parts; the spring element is not continuous nor is it compressively positioned between the grooves so as to occupy a generally kidney-shaped form. Also, the spring makes conformal contact with the grooves as opposed to edge contact as in Casler. The grooves have circular or elliptical profiles and do not have a V-shaped appearance.
. The friction polymers were attributed to the use of various organic and synthetic lubricants used to minimize friction and adhesion.
. Sperry also received companion patents, United States Patent No. 4,086,909, entitled "Electrical Contact Assembly and Method and Apparatus for Assembling the Same,” and United States Patent No. 4,141,139, entitled "Tool for Assembling Electrical Contact Assembly," on January 17, 1978, and February 27, 1979, respectively.
. For example, the connectors had to maintain the low-noise requirement in over 100 separate channels for 10 million revolutions during the six-year mission.
.Mechanical tests analyzed fatigue and vacuum life. The relationship between preload and resistance and the required twist stiffness of the roll ring were investigated. For example, different platings were researched and different gasses were allowed into the roll ring assembly to determine their effect at the flexure/groove interface. Late in the project, work was done on a lubricated roll ring system. Different types of lubricants were tried in different quantities and with different application systems.
. Two of the three figures involve straight sides (Figures 1 and 8) and one involves somewhat contoured sides (Figure 9). One of the figures has sides that intersect at a point (Figure 1), and the other two have contoured apexes (Figures 8 and 9). But each of the grooves has distinct sides that intersect at an apex.
. In Du Pont, the Court of Appeals for the Federal Circuit recently stated:
It is entirely proper to use the specification to interpret what the patentee meant by the word or phrase in the claim. See, e.g., Loctite Corp. v. Ultraseal Ltd., 781 F.2d 861, 867. But this is not to be confused with adding an extraneous limitation appearing in the specification, which is improper. By ‘extraneous,’ we mean a limitation read into a claim from the specification wholly apart from any need to interpret what the patentee meant by particular words or phrases in the claim. “Where a specification does not require a limitation, that limitation should not be read from the specification into the claims.' Specialty Composites v. Cabot Corp., 845 F.2d 981, 987 (Fed. Cir.1988) (emphasis in original).
(Citation omitted.)
. The specification nowhere specifically states that an acute angle is necessary to produce wedging. For example, the specification indicates that an acute angle is a necessary prerequisite for achieving the irreversible property of a wedge. The claimed invention clearly requires wedging but does not unambiguously indicate that achieving the irreversible property of a wedge is necessary. With respect to the presence of acute angles in the grooves shown in each of the figures, absent some indication to the contrary, claim terms should not be interpreted as limited to the figures in the specification. See, e.g., Lemelson, 752 F.2d at 1552, 224 USPQ at 534; Fromson, 720 F.2d at 1568, 219 USPQ at 1139; Strumskis v. United States, 200 Ct.Cl. 668, 675, 474 F.2d 623, 627, 175 USPQ 243, 245, cert. denied, 414 U.S. 1067, 94 S.Ct. 576, 38 L.Ed.2d 472 (1973).
. Of course, if as a matter of fact wedging will result only in situations where the angles are acute, interpreting the claim as requiring that the angle be small enough to result in wedging will produce the same result as if the claim were interpreted specifically to require grooves with acute angles.
. The Fjellstedt patent, entitled "Electrical Switch,” was issued on August 9, 1960.
. Both parties take the position that whenever the transfer contact is distorted into a generally kidney-shaped form, it will necessarily also display an appreciable arc of contact. Of course, as described above, the parties differ as to how much of an arc is necessary before the transfer contact is deemed distorted into a generally kidney-shaped form.
. Even where the claim literally reads on the accused device, however, infringement is avoided under the "reverse doctrine of equivalents” if the alleged infringer can show that the accused device "is so far changed in principle from a patented article that it performs the same or a similar function in a substantially different way.” Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 608-09, 70 S.Ct. 854, 856-57, 94 L.Ed. 1097, 85 USPQ 328, 330 (1950). See also SRI Int'l v. Matsushita Electric Corp. of Am., 775 F.2d 1107, 1123, 227 USPQ 577, 587 (Fed.Cir.1985).