OPINION
SETO, Judge.Plaintiff seeks recovery under 28 U.S.C. § 1498 of reasonable and entire compensation for the unauthorized use and manufacture, by or for the United States, of the invention described and claimed in U.S. Letters Patent No. 3,379,889 (“ ‘889” or “Barnett patent”), entitled “Beam-Driven Gyroscope Device”. The patent was issued on April 23, 1968, on an application filed April 5, 1963, by coinventor-plaintiffs Eugene R. Barnett, and his since deceased father, Willard L. Barnett. Plaintiff, Eugene R. Barnett, resides at 6268 Windsor Drive, Indianapolis, Indiana.
A trial on the merits was held specifically addressing the issues of validity and infringement of claim 7 of the patent in suit. The accounting phase was deferred for a later trial, pending the determination of liability.
Background Of The Barnett Patent
The application for the patent in suit was filed on April 5, 1963, by plaintiffs Eugene R. Barnett and his father Willard L. Barnett, the latter since deceased. The application described a gyroscope device in which the rotor was suspended by an elec-tro-magnetic field and light beams were used to fulfill several other necessary gyroscope functions. The device is best shown in Fig. 1 of the Barnett patent which is reproduced below to facilitate reference.
*635
The gyroscope consists of a generally circular rotor 10 mounted within a stator 12 attached to a housing 14. Electromagnets 22 set up an electromagnetic field which suspends the rotor in the center of the stator. The translational position of the rotor within the stator is monitored by a photoelectric system consisting of lights 28 projecting light beams 26 onto photocells 30. Movement of the rotor from the desired position alters the pattern of light falling on the photocells which send signals through control device 18 to regulate the electromagnets to return the rotor to the desired position. Another set of light beams 36 is created by lamp array 34 to cause the rotor to spin. This is accomplished by a set of recessed radial vanes 38 *636mounted along the equator of the rotor which intercept light beams 36 and provide a rotational force on the rotor through the radiometer effect. The radiometer effect is caused by the unequal rebound of the residual gas molecules in the stator off the dark side of the vanes which are heated by the energy of the light beams. In order to determine or to “pickoff” the information concerning the angular orientation of the rotor’s spin axis, a lamp 40 and focusing lens 42 are mounted at one end of the rotor’s spin axis to project light ray 43 onto photocell array 44 mounted on the stator. Lamp 40 is powered by an active energy conversion system mounted on the rotor, where radio meter vanes 38 also contain photovoltaic cells 46 which receive energy from rotor drive light beams 36 and convert this light energy into an electrical current, which flows through wires 48 to power the lamp. The orientation of the rotor’s spin axis is determined by the position at which light ray 43 falls upon the photocell array 44.
Background And Prior Art
The basic principle upon which gyroscopes are based, is that a rotating body tends to maintain a stable orientation in space. In the mid-nineteenth century, the French scientist J.B.L. Fouchault constructed a gyroscope in which the rotor was mounted in three supporting frames known as gimbals in which it moved freely, and demonstrated that a rotor maintained its original orientation with respect to inertial space, and did not follow the earth’s rotation. Fouchault was also responsible for naming the device a gyroscope. While Fouchault’s experiments suggested that the gyroscope could be used to indicate direction, the first practicable gyroscope compass was not developed until the early 1900’s when it was used for marine applications.
Until World War II, gyroscopes were usually gimballed devices in which the rotor was mechanically attached to its gimbals through bearings. However, the friction attributable to the bearings caused the rotor to precess or drift from the stable position. Precession is attributable to the physics of angular momentum whereby a torque acting at right angles to the spin axis of a rotating body causes the axis to move at a right angle to the force. Gyroscopes of the 1940’s used jeweled bearings, and typically had drift rates of the order of one to ten degrees per hour. The development of the ballistic missile by Germany near the end of World War II, and its improvement through the 1950’s into a missile having intercontinental range, as well as the adaptation of the nuclear submarine to serve as a mobile sea-based launching platform for ballistic missiles, created a need for gyroscopes with much lower drift rates, of the order of one-tenth or one-hundredth degree per hour, to be used in the inertial navigation systems of these missiles and submarines.
A basic principle of an inertial navigation system is the computation of a vehicle’s present position by monitoring acceleration forces acting on a navigation instrument carried by it. The navigation device is first stored with initial position information, and the changes in the position of the vehicle are constantly computed by performing a double integration with respect to time of the acceleration forces sensed by the instrument. Acceleration forces are usually sensed by three accelerometers mounted in orthogonal relationship upon a stable inertial platform. The platform is maintained in a stable relationship to space or to the surface of the earth by one or more gyroscopes.
In an effort to minimize the drift rate in gyroscopes, efforts were made to reduce gimbal friction or to eliminate the gimbal friction or to eliminate the gimbal bearings altogether. During the late 1940’s and the early 1950’s, floated gyroscopes were constructed whereby the gimbal assembly was “floated” in a liquid or supported by a current of air to reduce the load carried by the gimbals. The air or liquid-floated bearing principle was also applied to the rotor itself to eliminate the need for a mechanical bearing between the rotor and stator of the gyroscope.
*637In the early 1950’s, electrically suspended gyroscopes were developed by Arnold Nordsieck at the University of Illinois and Jesse Beams at the University of Virginia. Dr. Nordsieck’s device used an electrostatic field to suspend the gyroscope rotor, and is described in his U.S. Patent No. 3,003,356, which was filed on November 1, 1954. Jesse Beams’ gyroscope used an electromagnetic field for this purpose, and is described in U.S. Patent No. 2,691,306, filed on January 30, 1951. Hence, well before 1963, it was known in the art that a free rotor gyroscope could be constructed without gimbals or other mechanical bearings by suspending a gyroscope rotor with a fluid or gas bearing, or by utilizing an electrostatic or electromagnetic field.
As previously noted, the property which makes the gyroscope a useful device, is the tendency of the angular momentum of the spinning rotor to maintain the orientation of the spin axis of the rotor fixed in relation to inertial space. It is therefore necessary to be able to ascertain the angular orientation of the rotor’s spin axis, a function which is known in the gyroscope art as the rotor spin-axis pickoff scheme (“pickoff scheme”). In a free rotor gyroscope, in which mechanical bearings have been eliminated through the use of a fluid or gas bearing or an electric field to suspend the rotor, it is undesirable to require any mechanical connection to the rotor (from the stator) to determine the spin-axis orientation. This led to the concurrent development of optical and electrical systems for detecting changing rotor spin-axis orientation in free rotor gyroscopes. Nordsieck utilized the capacitance effect of a capacitive bridge, formed by electrodes on the stator and a protuberance on the rotor, to sense rotor orientation. Rotor-position pickoff schemes, utilizing the electromagnetic inductance effect, were also known in the art prior to 1963. Similar optical schemes, utilizing photoelectric pickoff sensing, were well known in the art by that time.
Vehicles utilizing gyroscopes include aircraft, ballistic missiles, surface ships and submarines. The accused gyroscopes are manufactured for defendant by the Rockwell International Corporation.
Relevant Definitions
A “gyroscope ” is a device in which one of its elements, the “rotor ”, rotates about an imaginary axis, generally referred to as the rotor’s “spin-axis ”, going through the rotor’s center of mass. That portion of the gyroscope which is immediately adjacent and exterior to the spinning rotor is generally referred to as the gyroscope’s “stator”.
The rotor’s positional “stability ” depends on its angular momentum, the product of its rotational moment of inertia and the angular velocity. Therefore, the greater the rotor’s inertia or its rotational speed, the more stable is its spin axis. When an external torque acts on the spinning rotor at an angle to its spin axis, the spin axis is deflected and tends to rotate about an axis normal to the external torque vector; this motion is called “precession ” of the rotor’s spin axis. The rotor’s performance, however, is independent of the direction in which it rotates.
A long-standing goal of gyroscope designers has been to support the rotor inside a stator in a way that obviates external torques acting on the rotor, and hence eliminating precession of the rotor’s spin axis. One obvious problem in early gyroscope designs was the frictional torque at the rotor bearings. As indicated supra, the solutions tried included liquid or gas floatation of the rotor, as well as electro-magnetic or electrostatic force fields strong enough to levitate or suspend the rotor, with no physical contact between it and the stator.
Both these suspension modes, i.e., electromagnetic and electrostatic, require motors made of electrical-conducting materials, in which eddy currents are generated as the rotor spins. These eddy currents then interact with the causative external magnetic fields, e.g., the Earth’s own magnetic field; tending to slow the rotor’s spin, thereby causing precession. Various tech-*638ñiques, such as Rockwell’s mumetal shield, have been tried to eliminate the influence of such extraneous magnetic fields and the resultant precession-causing torques. See infra, view A of Rockwell’s mumetal shield:
Even so, gyroscopes intended for prolonged use are designed so that energy can be supplied via an externally applied torque to counter the drag torques experienced by the rotor from residual air or gas friction, or due to the influence of extraneous magnetic fields. Depending on how this is effected, and on the rotor’s size and mass, this replenishment or energy may be done intermittently or continually.
The accuracy of a gyroscope is characterized by its “drift rate”, measured in degrees per hour, the rate at which its rotor’s spin axis drifts away from a particular orientation in the inertial reference frame under the action of extraneous torques. The smaller the drift rate is, the more stable the gyroscope. Drift rates of about 1 degree per hour were achieved by the end of World War II and current capabilities are better than 0.01 degree per hour. It is accepted practice to ensure a stable and measurable drift rate, since it cannot be ignored for accurate missions, and then to compensate for it. This may necessitate knowing, or controlling, the rotor’s angular speed. The amount of the angular error, once knowing the rotQr’s initial spin-axis position, depends on the time integral of the drift rate. An inertial navigation sys-*639tern can be considered to be a special case of a dead-reckoning navigation system, in which the errors grow with time, because of the drift-rate of the gyroscope. The user’s desire is to keep such errors as low as possible, particularly for long-range missions, e.g., in nucléar submarines on patrol, in order to avoid requiring the vessel to expose itself when obtaining an external position fix.
“Inertial Navigation” of the vehicles carrying the gyroscope, requires that the vehicle’s motion be purposeful in as many as all three spatial dimensions, and in time. In practice, this generally requires knowledge of where the vehicle started its motion, the magnitude and direction it is moving at any instant, and the magnitude and direction of its instantaneous acceleration at any moment. The gyroscope, with separate accelerometers (Rockwell), or by itself (Barnett), provides the instantaneous magnitude and direction of the vehicle’s acceleration vector as it moves. Knowing both the direction and magnitude of translational acceleration at each of two successive instants in time, the user will integrate the acceleration vector over time; once, to determine his instantaneous velocity vector; and then a second time to determine his instantaneous position, i.e., exactly where he is with respect to his starting point. Once knowing the desired direction, and the particular route, one can chart a course accordingly; monitor the progress; and make course corrections. In practice, this is done on a computer because of the large volume of data and calculations involved.
The gyroscope’s spinning rotor provides a reference direction, either in constant time, or one whose drift rate is known and compensated therefor. In addition, the magnitude of the vehicle’s translational acceleration must be measured in three mutually orthogonal directions, whose orientation with respect to the rotor’s spin axis is known. Some gyroscopes have so-called “stable platforms”, which orientation is kept fixed with respect to the stable rotor axis (or axes when more than one gyroscope is utilized simultaneously), to carry three mutually orthogonal accelerometers. The preferred embodiment discussed in the Barnett patent teaches an alternative way of measuring acceleration, both in magnitude and direction, in the normal course of operating the gyroscope, without separate accelerometers outside the gyroscope itself. However, claim 7, the only Barnett claim in issue, is not concerned with the specific method by which translational acceleration is measured.
The stator cavity is usually similar in shape to the rotor contained within. The Barnett rotor, when it spins, has an essentially spherical shape. The Rockwell rotor is spherical in shape but describes a slightly torodial path because its center of rotation does not coincide with its geometric center. Both the Barnett and Rockwell gyroscopes have spherical stator cavities. In normal operation, the rotor’s center of rotation (usually also its center of mass) is maintained at the stator’s geometric center regardless of the rotor’s shape. Any translational deviation from this position must be rapidly determined and promptly corrected. This “centering” of the rotor is done by means located on the stator capable of sensing the rotor’s spin-axis position, and actuating the translational corrective response, to return the rotor to its specified position by applying a carefully monitored force from the stator. This “suspension force” overcomes gravitational and translational accelerations experienced by the rotor. “Newton’s Second Law of Motion ” requires that the force applied by the stator to the free rotor be directly proportional to the relative acceleration between them. This statement puts into words the basic equation of Newtonian mechanics:
P = M x a
(force) = (mass) X (acceleration)
There are two ways of supporting the gyroscope’s stator in the user vehicle: first, by having it fixed rigidly to the vehicle (a strapdown gyroscope) or, second, by having it carried on one or more gimbals (a gimballed gyroscope).
(a) The “strapdown gyroscope” needs to have no moving parts other than its *640rotor and may, therefore, be relatively sturdy, simple and inexpensive. As the user vehicle moves about, the stator of the strapdown gyroscope moves with it vis-a-vis the rotor’s stable spin axis. Thus, the motion of the stator, measured with respect to a reference direction related to the rotor’s axial position, represents the motion of the vehicle itself.
(b) A “gimbal ” is probably best visualized in its simplest form to be like a universal joint with two bearing axes orthogonal to each other, such that a body, e.g., a gyroscope stator, can turn about the axes through those bearings. A complete set of such assemblies would provide for complete isolation of anything carried inside the gimbals from the external motion of the user vehicle. See picture, infra, showing a model of the electrostatically supported gyroscope monitor (ESGM) stable platform manufactured by Rockwell International, displaying gimbal axes, gimbal rings, and stable element.
For prolonged missions, e.g., on nuclear submarines that stay submerged for months at a time, a “gimballed gyroscope” supported so as to have four degrees of freedom (one being redundant), with servomotors controlling each gimbal, is preferred over the strapdown type. Deviations from the specified relationship between the stator and the rotor’s axial position, generally referred to as an “error function ” or “error signal ”, are utilized to manipulate the gimbal servos for corrective action. A gimballed gyroscope is capable of much greater accuracy than the strapdown type. The user actually measures the relative change in angular position between the stator and the vehicle, since the stator’s orientation is essentially locked on to that of the stable rotor spin axis.
*641
Electrostatically Supported Gyroscope Monitor (ESGM)
A gyroscope capable of functioning in the strapdown mode can be utilized in the gimballed mode when it is mounted on proper gimbals. The reverse is not true; a gyroscope designed so that the rotor and stator must maintain very closely a specified orientation with respect to each other cannot function as a strapdown gyroscope. The specific gyroscope embodiment disclosed in the Nordsieck patent cannot be used as a strapdown gyroscope. The Rockwell gyroscopes, when used in the gim-balled mode, are firmly affixed to a stable platform which carries three mutually orthogonal accelerometers to measure the three components of translational accelera*642tion in the inertial reference frame. The stable platform carrying the accelerometers, therefore, has its angular position unchanged in the inertial reference frame.
A “capacitor ” is any device that holds an electric charge in response to a voltage applied across two electrically conducting elements called “electrodes”. These electrodes must be separated by an electrically non-conducting material called an insulator or a “dielectric ”, e.g., either air or, preferably, a vacuum. The electric charge held by the capacitor is directly proportional to the capacitance times the voltage, i.e., Q = C x V. In terms of units, one “coulomb ” of charge is the amount held in a capacitor of one “farad” when one “volt” of voltage is applied across the capacitor electrodes. Given a certain, fixed voltage difference between the two conducting elements of a capacitor, if one moves them closer, one increases the capacitance and hence the charge carried, and vice versa.
If one were to apply a source of direct current, e.g., a battery across the electrodes of a capacitor, there would be a build-up of positive charge on the electrode connected to the battery’s positive terminal, and a corresponding equal but negative charge on the other electrode. While the charge transfer was taking place, there would be an actual flow of electric charge through the wires connecting the battery and the capacitor, i.e., an “actual current” would flow in the external circuit and light up a light bulb, if one were included in the current path, until the capacitor was fully charged. Therefore, no more actual current can flow through the external circuit when the capacitor is fully charged.
While an actual current was flowing through the external circuit in the preceding illustration, no electrons physically crossed the dielectric and yet, at the end, there is a definite charge separation across the capacitor plates (i.e., the two conducting elements) of the capacitor. Thus, a capacitor, when it has an externally imposed voltage on it, stores up electrical energy. The charge separation across the capacitor occurs due to the flow of an “apparent current” which does not involve the actual flow of charged particles across the dielectric. The charges merely appear to flow across and out of the capacitor. If an alternating current were used in the preceding illustration, i.e., a current varying cyclically in time, so long as there was an actual current in the external circuit, there would also be an apparent current of the same frequency across the capacitor. The capacitor stores up and discharges electrical charge in response to the externally imposed alternative voltage.
Such a capacitance of magnitude C, carrying a charge Q, possesses electrostatic capacitive “energy” in the amount Q2/2C, and can do this much “work ” in the process of being discharged. This capacitive energy is considered to reside in the electric field between the plates, although in fact, the charges of opposite signs, whose separation constitutes the stored energy, actually reside at the electrodes and not in the dielectric between them. Thus, a capacitor, when it has an external voltage applied across it, experiences a separation of charge across it by way of an apparent current, and thereby stores up electrical energy. This is true, whether the applied voltage is constant, or time-varying.
Electrical energy, like any other form of energy, is a conserved quantity, which can be neither created nor destroyed. It may, however, be exchanged among various bodies, or may be converted from one form to another. The existence of positive charge on one electrode of a capacitor, with an equal amount of negative charge on the other, causes the two elements to be attracted to each other, since opposite charges attract. If one of the electrodes is fixed in space and the other is free to move, then the movable one will try to move towards the fixed one. Such an attractive force can be used to overcome inertia forces to accelerate a free body, per Newton’s First Law of Motion, e.g., to recenter a free rotor in the ESG, when the stator undergoes translational acceleration.
*643 Prosecution History Estoppel
In the first Office Action dated April 7, 1964, the Patent Office Examiner cited four United States patents as prior art and rejected all eleven claims in the Barnett application. Wittkuhns, U.S. Patent No. 1,999,646, Annen, U.S. Patent No. 2,378,-744, and Dias, U.S. Patent No. 2,541,217, were each cited as disclosing the use of magnetic means to locate a gyroscope member in a stator with an energy source and conversion means to impart rotation to the gyroscope and sense its direction of movement. More specifically, Wittkuhns and Annen disclose gyroscopes in which the movement of the rotor’s spin axis is detected by a photoelectric system which maintains the desired orientation of the rotor within the stator. Dias discloses the use of a photoelectric system to determine the position of a compass needle in an automatic ship steering system. Hammond teaches the use of a photoelectric system mounted on the rotating member of a steering control system in which the rotating member tracks the position of a light source.
In an amendment filed October 7, 1964, applicants argued that the references cited by the Patent Office did not disclose the use of a magnetic means to locate the gyroscope member, as the Examiner had asserted, and that the reference also failed to disclose several specific features of the applicants’ gyroscope. Of particular relevance to this case, applicants stated in reference to application claim 8 (which eventually issued as patent claim 7 in suit), that the references failed to show the use of an energy conversion means located on the rotor which converts some of the light used to spin the rotor into electrical energy to power the light which projects the spin-axis light ray 43 from the rotor to the stator. In response to these arguments, the Examiner issued a second Office Action dated December 28, 1966, in which ten pri- or art patents were cited to reject all claims as obvious under 35 U.S.C. § 103. The principal reference was Parker, U.S. Patent No. 2,919,583 which also discloses a magnetically supported gyroscope.
In citing Parker, the Examiner explicated:
Parker shows a magnetically supported gyroscope and a photoelectric system for sensing and correcting the position of the rotor. [Moreover,] Parker provides electron beam means aimed tangentially of the rotor for driving the rotor without mechanical contact. [Emphasis supplied.]
Figure 1 of the Parker patent is reproduced below:
*644
The Parker gyroscope consists of a spherical rotor 10 mounted in an enclosure 25 and suspended by an electromagnetic field created by electromagnets n, 12, IB, and M. The translational position of the rotor is maintained by a photoelectric control system comprised of lamp IB and photoelectric cell 22 in essentially the same manner as the Barnett gyroscope. The rotor is driven by an electron gun 27 which projects a beam of electrons 28 tangentially on the surface of the rotor. The position of the rotor’s spin axis is determined by photoelectric detector 71 which detects the modulation pattern of the beam of light projected by lamp IB upon the rotor and reflected by a band having different optical *645properties placed on the surface of the rotor.
In this second Office Action, Beeh, U.S. Patent No. 3,268,735; Simon, U.S. Patent No. 3,225,608; Elwell, et al., U.S. Patent No. 3,254,537; and Marrison, U.S. Patent No. 2,919,358, were cited by the Examiner with the following assertions:
Beeh shows a radiometer drive means for rotating a light modulator which is analogous to Parker’s drive means. Simon and Elwell et al. show photoelectric means for sensing and correcting the axis of rotation of the rotor which are the equivalents of applicants’ means. The use of photovoltaic cells to generate power without mechanical connections to a generator is too well known to require additional support, but it is noted that Marrison shows this concept. Thus, it is held that prior art renders obvious the instant invention as defined by the presented claims. [Emphasis supplied.]
Thus, Marrison, U.S. Patent No. 2,919,-358, was cited as showing the use of photovoltaic cells to generate current within a rotor without mechanical connections.
Applicants responded to the second Office Action in an amendment filed March 27, 1967. In this amendment, the applicants explicitly distinguished their invention from the gyroscope shown in Parker, by focusing the Examiner’s attention to the double function performed by Barnett’s energy beams 36 which provides both the imparting rotation of the rotor, and the energizing of the rotor-circuit 48, which energizes energy-means 40, which projects energy ray 43, for the rotation-sensing cells 44. In contrast, the primary reference Parker, uses an electron beam 28 to impart rotation to its rotor 10, but does not use that electron beam for the second purpose as stated by the Barnett coinventors. On June 23,1967, the Examiner allowed claims 1-3, 5-10 and 12 “in view of applicants’ persuasive and clearly presented arguments.” Application claim 8 became claim 7 in the patent in suit which issued on April 23, 1968.
Teaching Of The Barnett Patent
The Barnett invention relates “particularly to control means utilizing gyroscope components for control of vehicles.” The coinventor-plaintiffs patented a “new and improved gyroscope control means having advantageous features of sensing, operation and responsiveness.” Without using language limiting the practice of their invention to any particular means, the plaintiffs explain in their patent specification, how cooperating elements of their illustrative embodiment would perform the three essential functions of a useful gyroscope.
The various concepts and features of the Barnett invention are presented in the patent specification under these headings:
(a) Means for locating the rotor in the stator, and for sensing and correcting location-deviation of the rotor;
(b) Means for imparting rotation to the rotor; and
(c) Means for sensing and correcting relative rotation of the rotor and stator.
The Barnett patent contains a “ball-like” essentially spherical free rotor which, in use, is suspended under the action of forces exerted on it from the essentially spherical stator surrounding it. The term “essentially spherical” means that while the shapes referred to possess overall sphericity of geometry, both the Barnett rotor and stator may have recesses within them to accommodate various elements to perform specific functions.
In the illustrative embodiment, the Barnett patent discloses electromagnets in the stator, capable of generating a composite electromagnetic force field within the stator cavity, to exert a net force on magnetic material in the rotor to suspend or levitate it as a free rotor, under all operating conditions. These are electromagnets identified as 22 in Figures 1, 2 and 4 of the Barnett patent. They are not permanent magnets each of a fixed strength but, by design, individual electromagnets along each geometric axis, exerting precise control from all directions to counter gravitational and translational accelerations to maintain the *646rotor at the center of the stator cavity. The Barnett patent shows a control unit IB, to exercise, inter alia, this suspension/location control over the electromagnets 22. There is nothing specific about the Barnett embodiment that requires only an electromagnetic suspension; an electrostatic drive could be utilized in the Barnett gyroscope, assuming the separation between the rotor and stator was made substantially smaller, and this achievement within one skilled in the art.1 The Barnett patent teaches that the energy required to center the rotor, may be used to determine the user vehicle’s translational acceleration.
In its illustrative embodiment, the Barnett patent suggests a multiplicity of narrow light beams 26, projected from sources 28, on the stator so as to very closely bracket the spinning rotor IB and reach light sensitive photoelectric cells 30 connected to the control unit 18. When the rotor moves, relative to its designed-een-tered position inside the stator, it blocks one or more of these light beams 26. This affects corresponding photoelectric cells 30 and provides error signals to control unit 18, thereby causing it to adjust the suspension forces exerted by suspension electromagnets 22. Thus, the Barnett patent shows in its illustrative embodiment a photoelectric rotor location system that works cooperatively with the electromagnetic suspension system to correct rotor-to-stator spin-axis deviation; while the latter system simultaneously measures the vehicle’s translational acceleration.
The Barnett patent, in its illustrative embodiment, does not suggest that the figures are drawn to scale, nor does it specify any particular magnitude for the stator or rotor geometries, rotor speed, number or size of rotor vanes, electrical power to the suspension electromagnets, or sensitivity for the photoelectric cells. Persons skilled in the art, seeking to apply its teaching, presumably would select such parameters to suit the projected use or mission at hand, and may readily adjust the number and direetion of rotor-locating light beams to minimize the radial gap between the rotor and stator. Likewise, for different purposes, e.g., rotor spin-axis location or rotor location sensing, light of different frequencies may be used to avoid interference among different light-sensitive elements as they function cooperatively.
Once the Barnett free rotor is levitated and centered, it must be spun up to a useful speed, i.e., it must be given sufficient angular momentum to provide a stable axial position for reference by the user. However, despite substantial evacuation of the stator cavity, there will remain some low pressure gas around the rotor, exerting a frictional drag which would in time slow it down to an unacceptably low speed. Since its rotor is likely to contain at least some electrically conducting material, there may also be eddy currents generated therein by extraneous magnetic fields, e.g., the earth’s own magnetic field, which may lead to interactions causing additional drag torques slowing down the rotor. Hence the rotor’s kinetic energy must be continually replaced by applying an external torque to it.
Recognizing that the rotor must be spun up, and its kinetic energy thereafter replenished to overcome drag torques, the Barnett patent, in its illustrative embodiment, suggests a solution that requires neither magnetic forces nor a mass transfer from the stator to the rotor, to achieve and maintain rotor spin. The suggested technique involves certain energy sources, i.e., lights 34 in the Barnett patent, Figures 1 and 5, projecting energy 36 from the stator 1.2 to the spinning rotor 10. The amount of energy so projected would be regulated by control unit IB. The rotor would have around its “equator”, i.e., in the plane normal to the spin axis, a set of vanes each colored dark on one side and light on the other. Persons skilled in the art could shape the vanes so as to reduce the gas frictional drag. See Figures 3 to 7, infra, of the Barnett patent:
*647
In one scenario, the dark face of each vane would absorb more of the incident energy, and be at a higher temperature than its opposite lighter-colored and more reflective side. When the residual gas molecules, still present after substantial evacuation of the stator cavity, bounce off the darker side of each vane, they do so at a higher velocity than they do from the lighter side. As a result, per Newton’s Third *648Law of Motion, as they leave the dark side of each vane, they impart to it a greater reactive momentum than they do on the light side. This disparity in the net force on each of many vanes, acting at a distance from the rotor’s spin axis, generates a torque capable of speeding up a rotor from rest, or of replenishing the rotor’s kinetic energy of rotation in spite of drag torques. This effect, generally referred to as the “radiometric effect”, requires the presence of some gas around the rotor. The Barnett rotor would turn by this means, even with air at atmospheric pressure around it, albeit slowly. However, one skilled in the art may consider using low pressures and various types of air molecules of low molecular weight, e.g., hydrogen, to increase rotor spin speed significantly. The darker vanes in this scenario will always move away from the incident light, thus deciding the direction of the rotor’s spin.
In an alternative scenario, even if all the gas molecules were removed from the stator cavity, one skilled in the art could still cause the Barnett free rotor to rotate sufficiently to function as a gyroscope. The photon theory of light, which is an alternative to the wave-theory of light, provides the explanation for such a rotation, as follows: photons are discrete quanta of light energy, which reach both light-colored and dark vane surfaces with equal incident momentum, and are reflected back with essentially reversed momentum at the lighter vane surfaces, and are simply absorbed at the darker surfaces. The consequence is that, for each incident photon, the lighter surface, reflecting it back reverses the photon’s momentum, and therefore itself experiences an impulsive reaction force, per the law of physics, equal to twice that experienced, for each photon received and absorbed, at that vane’s darker surface. The fact that the dark surface absorbs more energy — makes it hotter than the light surface; but since it is assumed that no gas is present in this scenario, there is no gas friction drag torque to impede the rotational motion of the rotor. Overall, given uniform lighting around the rotor, in this scenario, the rotor will turn so that the lighter surfaces move away from light incident on them. This is the exact opposite of what happens to the direction of rotation when the rebound of gas molecules spins the rotor.
Persons skilled in the art of gyroscope design, could reasonably be expected to consider both alternatives. In practice, it would be impossible to extract every single gas molecule from the stator cavity. Nevertheless, when a strong enough vacuum is obtained, i.e., the number of gas molecules is sufficiently reduced, the operation of the rotor would shift from scenario one to scenario two; and the Barnett gyroscope would work equally well with the rotor spinning in either direction. The incident light energy does not have to have a preferred orientation either; uniform illumination around the stator will work perfectly well and, in fact, would be necessary for strap-down gyroscope operation.
Energy inputs to the stator, no matter how caused, would heat up the gyroscope unless proper cooling is provided. The Barnett specification allows for temperature control in element 20, regulated by control unit 18. Since there is no disclosure regarding element 20, the patentee apparently assumes it to be within the technology of one skilled in the gyroscope art.
In its illustrative embodiment, the Barnett patent discloses photovoltaic cells 46 integral with the rotor and distributed among the radiometric torquing vanes 38 at the equator of the rotor, wire 48 to carry the electrical current generated by these cells to a light bulb 40, the light therefrom being focused by lens 42, as a fine light beam 43 along the rotor’s spin axis, to receptive photo-sensitive cells 44, distributed about the stator’s inner surface. As the stator turns vis-a-vis the stable spinning rotor, the axially-directed light beam reaches different photocells 44. Control unit 18 responds to the signal from whichever individual photocell 44 is receiving the axially directed beam 43, and translates that photocells’ identity into axial-position information for reference by the user.
*649While the preferred embodiment of the Barnett gyroscope discloses the use of just one light beam projected axially from the rotor, one skilled in the art should recognize quickly, and without undue experimentation, that two light beams would be necessary to resolve the problem of determining the difference between: (1) when the rotor merely translated with respect to the stator; and (2) when the rotor’s spin axis axially rotated or precessed with respect to the stator.
For fine resolution of angular pickoff of the rotor’s spin axis, in the preferred embodiment of the Barnett patent, it would be necessary to have not only very small receptor elements at the stator but equally important, one or more very fine light beams projected axially from the rotor.
The Rockwell ESG Gyroscope
The accused devices are known as the Electrostatically Supported Gyro Monitor (“ESGM”), which is used in the inertial navigation system of ballistic missile submarines, and the Electrostatically Suspended Gyro Navigator (“ESGN”) AN/WSN-3(v)2, which is used in the inertial navigation system of nuclear attack submarines. These gyroscopes are manufactured for the Navy, by the Autonetics Marine Systems Division of Rockwell International at Anaheim, California.
The predecessor of today’s Rockwell International, has long been involved in the development of precision gyroscopes and inertial navigation systems. Shortly after World War II in 1948, North American Aviation was one of several defense contractors which received technology from the German missile program, and began development of a line of missiles, as well as the inertial navigation systems for such missiles. North American Aviation merged with Rockwell Corporation in 1967, to form North American Rockwell, and eventually, Rockwell International. Rockwell has provided inertial navigation systems for the Minuteman intercontinental ballistic missile, as well as for all Polaris, Poseidon and Trident fleet ballistic missile submarines.
Rockwell began to investigate electro-statically supported gyroscopes (“ESG’s”) in 1959. In 1968, a miniature ESG was constructed for use in airborne applications. This led to the development in 1971, of an ESG for submarine applications. Following competition with a Honeywell ESG, the Rockwell ESG was selected by the Navy for installation on ballistic missile and attack submarines, in 1974 and 1975, respectively.
The present form of the accused gyroscopes, also called the micro-ESG, was conceived in 1966, and first built in 1968. It contains a solid beryllium rotor one centimeter in diameter. In principle, however, an ESG rotor needs only to have an electrically conducting surface, and may have an interior made of electrically insulating material.
While the ESGM and ESGN systems vary somewhat in physical arrangement and electrical circuitry, both use two elec-trostatically supported gyroscopes to stabilize the inertial platform on which the accelerometers are mounted. That which is most relevant to this litigation, the gyroscope spinner assembly, which consists of the gyroscope rotor and its housing, is identical in the ESG’s used in the two systems. The component parts of the spinner assembly were shown before, supra. A replica of a demonstrative exhibit used at trial to explain the structure of the Rockwell gyroscope spinner assembly, was submitted for the record as DX-46A and 46B.
The rotor of the Rockwell ESG is a solid beryllium sphere, one centimeter (about one-half inch) in diameter. The sphere is embedded with two tantalum wires, set off from the center of the rotor, as shown in U.S. Patent No. 3,880,606. The rotor is surrounded by four pairs of diametrically opposed octantal electrodes mounted on the upper and lower hemispheres of the envelope. The width of the gap between the rotor and the electrodes is approximately 300 millionths of an inch (300 microinches). The octantal electrodes are connected to four electrical suspension circuits, which supply a 20,000 Hertz (20 kHz) 300 volt *650alternating current to the electrodes. It produces an electrostatic field strength of approximately one million volts per inch within the gap, which causes the spherical rotor to levitate to the center of the cavity. A set of induction torquing coils surrounds the rotor assembly and produces a rotating electromagnetic field, to initially orient the rotor and bring it to its operational spin speed at 3600 rps (revolutions per second). Once operational spin speed is achieved, the induction torquing coils are deenergized, and the rotor continues to spin in the evacuated assembly. The induction torquing coils are not energized when the gyroscope is operating in the normal navigation mode.
The rotor interacts with the octantal electrodes in the manner of a capacitive bridge. Each electrode and the adjacent surface of the rotor, forms a capacitor, in which electrical energy is stored in the electrostatic suspension field between the electrode and the rotor. If forces acting on the device cause the translational position of the rotor to stray from the center of the stator cavity, the capacitive bridge becomes unbalanced, and the external suspension circuitry applies corrective voltages to the electrodes, to return the rotor to the desired position in the center of the stator.
A perfectly uniform sphere will rotate so that its geometric center coincides with its center of mass. Since such perfection is impossible to obtain in practice, there will always be some displacement of one from the other. Rockwell decided to intentionally introduce a known amount of mass unbalance at a preselected position by adding tantalum, to an otherwise solid beryllium rotor. As a result, the ESG rotor does not spin about its geometric center, but rather, at or very close to its mass center. The consequence is that the geometric center’s path describes a plane normal to the spin axis, thus providing a plane of reference perpendicular to, and indicative of, the orientation of the spin-axis of the rotor. In use, the rotor’s surface sweeps out a toro-dial space, and alternately approaches and recedes from various points on the stator at its rotational frequency, approximately 3.6 kHz, i.e., 3600 cycles per second.
The ESG rotor is a solid beryllium sphere, 1 cm. in diameter, to which a small amount of tantalum is added away from the rotor’s geometric center. The added tantalum in early ESG models was “sputtered” onto one-half of the rotor’s exterior, and in later models was imbedded as two or three fine wires inside the rotor itself by means of a Rockwell patented process. Rotors with externally “sputtered-on” tantalum were not truly spherical, i.e., they had a “bump” on one side, and have not been used successfully.
Rockwell rotors with embedded tantalum wires are carefully machined to a spherical shape at about normal room temperatures. At the higher operating temperatures, with a high rotational speed, there is some balancing out of the anisotropic tendency of the beryllium to “prolate ” and the centrifugal effect to “oblate ” it, thus causing it to be spherical in use. This sphericity in use, is vitally important since the rotor must rotate about its principal axis for stability, and the electrostatic forces between the stator charges and the induced charges on the rotor surface, are always normal to the rotor surface.
Since the center of mass of the rotor is slightly offset from its geometric center as a result of the tantalum wires, the rotor rotates about its center of mass, rather than the geometric center, and the surface “wobbles” very slightly at the frequency of rotation. The amount of wobble is less than 25 microinches, i.e., less than one-tenth of the width of the very thin rotor-stator gap. This is smaller than the thickness of a sheet of paper, and cannot be discerned by the unaided human eye. The electrical effect of the very slight wobble is to redistribute the energy stored in the capacitive electrostatic suspension voltage at the 3600 Hz rotor spin frequency. This anisotropic phenomenon is identified as a mass unbalanced modulation (“MUM”). The magnitude and phase of the MUM signal on each of the four electrode plate pairs, is indicative of the angular relationship between the spin axis of the rotor, and *651a line drawn through the plate pair. External circuitry compares the magnitude and phase relationship of the MUM signals, from different plate pairs, and thereby determines the origination of the rotor spin axis with respect to the stator. This information constitutes the rotor-position pick-off signal, which is then used to drive the inertial platform’s gimbaltorque motors, to maintain the inertial platform in a stable relationship with respect to inertial space. The rotor-position pickoff concept is described further in Rockwell’s U.S. Patent No. 3,847,026.
As previously noted, the rotor is brought to operating speed by induction torquing coils, which are deenergized during normal operation. Since the stator envelope is evacuated to a very high vacuum, and the spinner assembly is shielded from the earth’s magnetic field by a mumetal enclosure; the rotor can therefore spin for many days before it would coast to a stop due to the combined friction of the remaining gas molecules, and the electromagnetic braking effect from the residual magnetism, which penetrates the mumetal shield. However, since the electrostatic force vector which levitates the rotor is directed toward the geometric center of the ball, a moment arm exists between the force, and the offset mass center of the ball about which it rotates. This provides a small accelerating torque to the rotor, which offsets the decelerating torque, attributable to gas and magnetic field drag. The suspension circuitry is provided with a speed controller, which creates a sharply tuned phase lag between the MUM pickoff signal and the electrostatic force applied to the rotor. The tuned circuit causes the rotation of the ball to be maintained at the desired speed. This concept is described in Rockwell’s U.S. Patent No. 3,906,804.
Immediately around the ESG rotor, and approximately 300 micro-inches larger in radius, is a spherical stator envelope, consisting of two mating hemispheres, plated inside with a smooth electrically conducting surface, which is machined to create eight “octantal stator electrodes ” controlled by an external circuit. Despite every attempt by Rockwell to make the eight octantal electrodes identical with each other, they are not precisely the same, because machines have tolerances. Thus, from one stator electrode to another, within a single ESG, and presumably between any two ESG’s, there are bound to be differences in individual electrode shape, size, thickness, surface properties and edge smoothness. In use, therefore, the electrostatic field supporting and centering the rotor, cannot be made truly homogeneous despite specification of very close tolerances. This small drag due to these stray electrostatic currents on the Rockwell ESG rotor, was never measured by Rockwell, nor was this asserted problem addressed by Rockwell in its patent for “Speed Control For An Elec-trostatically Supported Ball Gyroscope”, apparently indicating that the small drag was not serious enough to impede the proper functioning of the Rockwell gyroscope.
A variety of factors can induce electrostatic currents in rotors of the Rockwell type. Such factors as inhomogeneity in the stator, or inhomogeneity in the interior of the rotor, can all introduce small currents in the rotor surface which interact with the stator, and tend to provide slowing-down friction that decelerates the rotor.
In the ESG, the rotor and the eight stator electrodes are all made of electrically-conducting material separated from the rotor by a strong vacuum, with external voltage differences applied to diametrically opposed stator octants by pairs. The two diametrically opposed electrodes in any one of these four pairs, have charges of opposite sign, i.e., one has positive and the other a negative potential, with respect to the ground, for the external circuit connected to both.
At the very start, when the ESG stator is about to be powered-up, there are no voltages applied to any of the stator electrodes, and the spherical rotor surrounded by vacuum, is resting in physical contact with the octant immediately under it. If a voltage difference is now imposed across the pair of opposite octants (on one of which the *652rotor is resting) a voltage difference is built up between the rotor and the octantal electrode above it, across the vacuum between them. There is now a capacitor in existence, with the upper octantal stator electrode as one of its electrodes, and the rotor surface immediately adjacent to and under it across the vacuum, as the other electrode. Each carries opposite charges that attract each other. This force is proportional to the charge in the plate surface times the electric field between the two surfaces. By a judicious arrangement of capacitors above and below the rotor, and by a circuit that applies the right voltage to these capacitors, it is possible to supply the net force required to hold a rotor vertically suspended, against the force of gravity. It is in this manner that the ESG rotor is first levitated. Once it is levitated, it loses physical contact with the stator electrode on which it was initially resting. Therefore, there are now two capacitors, one at the top (between the top stator electrode and the adjacent portion of the rotor surface), and one below (between the stator electrode underneath the rotor and the adjacent portion of the rotor surface), with attractive forces being exerted on the rotor across a 300 microinch vacuum gap by each member of the four pair of stator electrodes. Since these two stator electrodes have charges of opposite signs on them at any instant, the corresponding charges induced on the surface of the rotor adjacent to each, will have charges of signs opposite to these. The rotor, therefore, has no net gain or deficit of electrical charge, and is itself at “ground potential” and represents a “virtual ground ”.
In the ESG circuit, one externally applied current applied to the four pairs of octantal electrodes, is alternating at 20 kHz, i.e., goes from peak “plus” value to peak “minus” value, with respect to ground, and back, 20,000 times per second. The external circuit, through this and other currents at other frequencies, controls the net current to each pair of electrodes to provide the exact amount of force required at any instant to keep the rotor centered against the effects of both gravity and of any translational accelerations experienced by the vehicle carrying the stator. Each electrode pair’s voltage is at a different phase, such that all four acting cooperatively provide the net force needed to keep the rotor centered at all times. A change in gap size will cause an alteration of the currents to octants, and thus activating corrective forces to reeenter the rotor.
Considering only the location and suspension functions of the electrostatic forces exerted by the stator on the rotor, it is obvious that, so far as the free rotor is concerned, there is acting on it a 20 kHz (20,000 cycles per second) basic frequency time-varying voltage from each pair of electrodes. The induced charge distribution on the rotor’s surface, therefore undergoes a redistribution of plus and minus charges in a complex pattern at 20 kHz (20,000 cycles per second), in response to these externally imposed voltages, even when the rotor is simply centered at rest in the stator, i.e., when the rotor is not spinning and before consideration is given to how the axial rotor-position pickoff function is accomplished. When the rotor interior is at a single ground potential, no charges flow through it; all charge redistribution must take place on the rotor by way of surface charges. In describing a hypothetical gyroscope with a rotor-stator relationship similar to the accused device, Dr. Allen hypothesized the existence of electrostat-ically induced charges being present on the rotor, which cause resistance-heat effects as these charges are redistributed on the rotor due to the alternating current nature of the 20 kHz suspension voltage, and the fluctuating rotor-stator capacitance caused by the mass-unbalanced rotation of the rotor. However, Dr. Allen testified that the heat created by the circulating currents on the rotor, does not provide information concerning the rotor spin-axis position, and is not directly involved in conveying the pick-off information.2 Moreover, there was no *653testimony from any expert at trial averring that the heat created a problem in the functioning of the accused Rockwell gyroscope. Absent testimony to the effect that the heat on the stator presented a sufficient problem in the accused Rockwell device, it cannot be established that the stator in the accused Rockwell device must be cooled down.
In actual use, once the rotor is levitated, it must be spun up to its operational speed. In the ESG, this is done by applying a rotating magnetic field, that first induces eddy currents in the rotor body, and, simultaneously interacts with the electromagnetic field generated by the rotor eddy currents, to provide an accelerating torque to spin up the rotor from rest to a typical operating speed of 216,000 r.p.m. The rotating magnetic field “motor” is then turned off, and a part of the external circuit, called the “servo circuit ”, takes control of the rotor’s motion, and either speeds it up or slows it down to the desired speed, and then maintains it at the proper value by continuously supplying energy, applying an input torque to overcome assorted drag torques that otherwise would slow down the rotor.
The rotor is now levitated and, under the combined action of the bias voltage of 20 kHz and the servo voltage, i.e., the rotation is such that the rotor’s geometric center is rotating about its spin axis. The surface of the rotor now approaches and recedes from the stator electrodes arrayed about the plane of rotation of the rotor’s geometric center at 3.6 kHz, thereby changing the cavity gap at any point at this cyclic rate. The voltage and charge on each of the rotor suspension plates is, therefore, affected at this 3.6 kHz cyclic rate across the capacitive-bridge-type circuit and is the origin of the MUM signals used through the ESG. Whatever the magnitude of this cyclic net voltage is at any electrode, it does not change until there is translational acceleration of the stator, or until the relative angular orientation of the rotor spin-axis position changes with respect to the stator. Any such change can be utilized to center the rotor, or to determine the change in the angular position of the rotor vis-a-vis the stator, i.e., to generate rotor spin-axis pickoff information.
The electrostatic force between each oc-tantal stator electrode and the adjacent rotor surface, acts normal to the rotor’s surface. The net stator force on the rotor therefore, acts through the rotor’s geometric center. If the rotor were a perfectly balanced sphere then its center of mass (and hence the location of its spin axis) would coincide with its geometric‘center, and the only consequence of a next external force would be to “center” the rotor. The ESG rotor, however, creates a situation in which the net force acting through the rotor’s geometric center does not necessarily also pass through the center of mass or the spin axis. There may, in fact, be a short distance serving as a “lever arm ” or “moment arm ” by which the net force on the stator misses the spin axis and exerts a torque on the rotor. This torque can either speed up or slow down the rotor, as may be required, when the electromagnetic spinup field cuts off. Additionally, this torque is used to replenish the losses in kinetic energy of rotation of the rotor by adding energy into the kinetic energy of rotation. No direct tests have been run to determine the drag torques acting on the rotor in the accused Rockwell gyroscope, because such tests were not feasible. However, tests were run by Rockwell on the accused gyroscope, to observe the change in speed of the rotor, or the change in the rotor drift rate, caused by altering factors which set up drag torques. Since the angular momentum of the rotor was known, the magnitude of the drag torques could be calculated, based on the magnitude of the changes in speed or drift rate caused by controlled alteration of the factors causing the drag torques.
*654The ESG rotor’s axial position (stable in the inertial reference frame or moving at a known drift rate), is determined with respect to the moving user vehicle, through an analysis of the voltages at the stator octants by the circuit controlling them. The operation of the suspension and rotor-position pickoff subsystems in the Rockwell ESG are explained in the Boltinghouse patent entitled, “Cross Product Pickoff For Ball Gyros Of The Electrostatic Levitation Type”, U.S. Patent No. 3,847,026 (“ ‘026”), and is applicable to the Rockwell ESGM and ESGN with respect to Figures 1, 2, 4 and 6a-6d of the Boltinghouse patent, see infra, and the relevant explanation provided in its patent specification. Its suspension electrodes are powered by the 20 kHz bias-current generator 17 which provides a constant bias current I<>. When the spherical rotor departs from the center of the cavity, it produces a voltage across transformer winding 23, which is called the rotor-position pickoff signal. The rotor’s axial position is then derived from analysis of the pickoff voltages.
*655
*656
*657
If a strapdown gyroscope is involved, then, as the vehicle turns with respect to the mass unbalanced rotor whose geometric center is rotating in a “reference plane ” normal to the spin axis, the relative position of each stator electrode will change with respect to this reference plane. As a result, the closest points of approach of the rotor with respect to each member of any pair of diametrically opposed electrodes, will shift. The proximity of the rotor surface adjacent to each stator octant, determines the charge on each element of the capacitor so formed. This charge on each octant must be modulated by the servo current, to ensure that the precise amount of force is exerted by that stator octant to maintain the rotor in its correct centered position. The angular orientation of the rotor vis-a-vis the stator, is determined from a comparison of the phase relationships of the various mass unbalanced modulation (MUM) pickoff signals, from the four electrode pairs.
In the alternative, if the gyroscope is being used in the gimballed mode, the alterations in servo currents to the stator electrodes serve as error signals to control the servomotors at each gimbal bearing. The servomotors act to eliminate these error signals by reorienting the stator so that it is again in congruence with the rotor in the latter’s new position relative to the user vehicle. Knowledge of the amount and direction in which the gimballed stator moved with respect to the user vehicle constitutes the rotor-position pickoff for navigation purposes. This may be used to stabilize a platform carrying three mutually orthogonal accelerometers to determine the user vehicle’s instantaneous translation acceleration, in the inertial reference frame.
Given two conducting elements forming the electrodes of a capacitor, i.e., one octant of the stator and in the rotor acting as the extension of the opposing octant of the stator, any relative motion between them will affect both the amount of charge held on each electrode and also the amount of energy, i.e., capacitance, stored in the capacitor, provided the voltage difference across the electrodes is maintained constant. In either case, the force between the two capacitor electrodes will change if the gap changes, while voltage is maintained constant. If, in addition, the voltage difference across the electrodes is varying at 20 kHz while the gap is changing the charge of the capacitor at a rate of 3.6 kHz *658(3600 times per second); then there will be a combined and complex time-dependent variation in the amount of charge at each electrode, in the amount of energy in the eight capacitors at any instant, and in the force between each electrode and the rotor. This scenario describes, in summary, what takes place between the individual stator octant and the adjacent surface of the rotor in the ESG, while the stator and the rotor are in a fixed angular orientation with respect to each other.
The servo current regulates the amount of charge at each octant in response to the rotor’s position adjacent to it, while corresponding charges of opposite sign move about on the rotor surface, and some of the energy is theoretically dissipated into heat, which could possibly increase the drag on the rotor by energizing the residual gas in the rotor-stator gap. The same energy source which is providing the servo current, puts this electrical energy back into the electric fields between the various oc-tants, and adds kinetic energy by continually torquing the rotor to maintain its rotational speed and hence the ESG operation. At a given time, an energy flow exists in the accused device between the stator and the rotor. This flow is the electrical energy which flows from one of the eight stator electrodes, which constitutes one electrode of the “electrode/rotor and electrode” capacitor, across the rotor-stator gap, to and through the rotor, which is instantaneously functioning as an extension of the opposing electrode/stator octant.
In the Rockwell ESG, it is necessary to replenish the losses in kinetic energy of rotation of the rotor, by providing energy to sustain the kinetic energy of rotation of the rotor. The rotor’s loss of kinetic energy is due to:
(1) residual gas in the rotor-stator cavity; and
(2) induced currents in the rotor which act on stray magnetic fields.
It is impossible, through the Rockwell ESG Mumetal Shield, to shield the Rockwell rotor from all of the earth’s magnetic forces. The residual gas, in the Rockwell rotor-stator gap, is heated somewhat by the heat from the Rockwell rotor. The Rockwell rotor apparently generates a small quantity of heat when its surface charges, which are being redistributed about the rotor 3600 times a second, meet the albeit small resistance of the metal of the rotor itself.
If the rotor were a perfectly balanced uniform sphere, its rotation would not modulate the charge distributions and voltages at the stator electrodes. But by having a deliberately unbalanced rotor, i.e., by using the MUM principle, the ESG generates voltage modulation signals whose magnitude and phase, by design, can be related to the reference plane in which the rotor’s geometric center rotates about is spin axis. The reason that the stator electrodes experience a voltage modulation, is that the charge on each stator electrode and its adjacent rotor surface, is related to the geometry of the gap between them, i.e., the gap size and where, say, the center of a particular electrode is located with respect to the rotor’s geometric center. As the rotor changes its angular orientation vis-a-vis the stator electrodes, its interaction with each individual electrode also changes. Therefore, the contribution of each electrode to the electrostatic field energy changes, and, simultaneously, the physical movement of the rotor repositioning the charges vis-a-vis the different electrodes modulate the voltages at those electrodes. This voltage modulation at the octants is the information flow that translates in the circuit to axial-pickoff information. Generating it requires the rotor-held charges to move around in different surface currents on the rotor, using up, through the generation of heat, some of the energy that is carried on the rotor. For continued use of the ESG, the flow of energy from the stator to the rotor must be maintained, i.e., the rotor’s energy must be continually replenished, for the user to continue receiving knowledge about its axial position. However, Dr. Allen was unable to determine how much energy loss occurred.
There is a storage of electrical energy, i.e., capacitance, in the eight capacitors *659formed by the octantal electrodes of the stator, the eight rotor surfaces directly adjacent thereto, and the gap between each therein. The transfer of energy in and out of these capacitors does not constitute a transfer of energy in and out of the rotor itself. Rather, the energy of each capacitor is stored in the gap between each octan-tal electrode of the stator and its respective adjacent rotor surface. Any change in the physical relationship (relative angular orientation) of one of the eight given capacitors, i.e., one of the single electrode-octants of the stator and its adjacent rotor surfaces, causes either an increase or decrease in the capacitance of the particular capacitor. As a particular octant of a rotor moves closer to a particular octantal electrode of the stator, the capacitance, i.e., the electrostatic energy in that rotor-stator gap, increases. If that same gap widens, the electrostatic energy in that gap decreases. It is this change (modulation) of the capacitance (electrostatic energy in each rotor-stator gap) within each of the eight rotor-stator gaps (gaps between each single electrode-octants of the stator and its adjacent rotor surface) that provides the information to the computer, which then discloses the new physical orientation of the rotor within the stator in the Rockwell ESG (also called the rotor-position pickoff scheme). During the spinning of the rotor within the stator, at approximately 3600 rpm, there is a rapid change in the electrical charges on the surface of each stator-electrode octant and its adjacent rotor surface, for each single revolution or spin. This rapid change or redistribution of charges on the respective surfaces of each stator-electrode octant and its adjacent rotor surface, causes a small loss of energy in the form of heat in the rotor and the stator. This loss of heat, which both parties agree is very small, has apparently never been measured and, therefore, inferred to be insignificant.
Infringement
Literal
We first address the issue of infringement, because the preponderance of the government’s case was directed toward proving non-infringement of claim 7 by the accused devices, rather than the invalidity of the claims of the Barnett patent.
The law of infringement requires that the claim or claims asserted be compared with and read on the devices or processes accused of infringement. Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 70 S.Ct. 854, 94 L.Ed. 1097, 85 U.S.P.Q. 328 (1950); Rel-Reeves, Inc. v. United States, 534 F.2d 274, 209 Ct.Cl. 595 (1976); Astra-Sjuco, A.B. v. U.S. International Trade Commission, 629 F.2d 682, 207 U.S.P.Q. 1, 67 CCPA 128 (1980); Amstar Corporation v. Envirotech Corporation and Energy Corporation and Energy Fuels Nuclear, Inc., 730 F.2d 1476, 221 U.S.P.Q. 649 (Fed.Cir.1984). There are two steps in determining patent infringement: (1) the subject matter of the claim or claims must be determined from a study of all relevant patent documents; and (2) the claims must be applied to the accused structures. Caterpillar Tractor Co. v. Berco S.P.A., 714 F.2d 1110, 219 U.S.P.Q. 185 (Fed.Cir.1983); Autogiro Co. of America v. United States, 384 F.2d 391, 401, 155 U.S.P.Q. 697, 705, 181 Ct.Cl. 55 (1962). Moreover, resort to the prosecution history, as well as the specification disclosure, is always appropriate in construing the scope of the claims at issue, regardless of whether the claims were amended during prosecution. Moreover, while an estoppel may be applied only as a result of amendments or arguments directed at establishing patentability over the prior art, the entire prosecution history may be used to determine construction of the claim language. McGill, Inc. v. John Zink Co., 736 F.2d 666, 221 U.S.P.Q. 944 (Fed.Cir.1984); SSIH Equipment, S.A. v. U.S. International Trade Commission, 718 F.2d 365, 218 U.S.P.Q. 678 (Fed.Cir.1983); Fromson v. Advance Offset Plate, Inc., 720 F.2d 1565, 219 U.S.P.Q. 1137 (Fed.Cir.1983). In addition, infringement is not to be determined by a comparison between parts of the description of a patent and the accused device, nor by a comparison between the *660accused and the patented device, Amstar Corporation v. Envirotech Corporation and Energy Fuels Nuclear, Inc., 730 F.2d 1476, 221 U.S.P.Q. 649 (Fed.Cir.1984); ACS Hospital Systems, Inc. v. Montefiore Hospital, 732 F.2d 1572, 221 U.S.P.Q. 929 (Fed.Cir.1984); McGill, Inc. v. John Zink Co., 736 F.2d 666, 221 U.S.P.Q. 944 (Fed.Cir.1984), and claims should be construed, if at all possible, to sustain their validity. ACS Hospital Systems, Inc. v. Montefiore Hospital, supra; Carman Industries, Inc. v. Wahl, 724 F.2d 932, 937 n. 5, 220 U.S.P.Q. 481, 485 n. 5 (Fed.Cir.1983); Klein v. Russell, 86 (19 Wall.) U.S. 433, 466, 22 L.Ed. 116 (1874). Finally, where a claim sets forth a “means” for performing a specific function, without describing any specific structure for performing that function, e.g., claim 7 of the patent in suit, the structures disclosed in the specification must be considered. Radio Steel & Mfg. Co. v. MTD Products, Inc., 731 F.2d 840, 221 U.S.P.Q. 657 (Fed.Cir.1984); Lockheed Aircraft Corp. v. United States, 553 F.2d 69, 82, 193 U.S.P.Q. 449, 460, 213 Ct.Cl. 395 (1977); Decca Ltd. v. United States, 420 F.2d 1010, 1014, 164 U.S.P.Q. 348, 351, 190 Ct.Cl. 454, cert. denied, 400 U.S. 865, 91 S.Ct. 102, 27 L.Ed.2d 104 (1970).
In the present case, plaintiffs are litigating only claim 7 of the Barnett patent, the accounting issue being deferred for a later trial, pending determination of liability. Only the issues of validity and infringement of claim 7 of the Barnett patent are before the court at this time, and we address infringement first.
Claim 7, presented in subparagraph form, reads as follows:
(a) A control means, comprising:
(b) a stator means
(c) and a movable means;
(d) locating means for locating said movable means in a certain location with respect to said stator means;
(e) first energy means, responsive to movement of said movable means away from said certain location, for energizing said locating means;
(f) second energy means for imparting rotation to said movable means; the said movable means being rotatable in response to energy from a second energy means;
(g) third energy means carried by the movable means; the said third energy means being actuated by said second energy means;
(h) indicating means carried by the stator means, responsive to energy of said third energy means, to indicate relative movement of the axis of the movable means with respect to the stator means;
(Claim element (f) is rearranged from the order in which its two parts appear in the patent.)
Claim 7 of the patent in suit is written in the means functions language as permitted by 35 U.S.C. § 112, which, in pertinent part, reads:
An element in a claim for a combination may be expressed as a means for step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof. [Emphasis supplied.]
Therefore, in order to determine the scope of claim 7, we must delineate and delimit the “structures” described in the Barnett specification. When we do so, it is clear that the following structures, which correspond to the claim’s means-for elements, are encompassed by claim 7:
Element in Claim 7 Corresponding Structure in Barnett Specification
stator means (element “b”) : stator 12 in Figure 1 of Barnett patent DX-41, page 3;
movable means (element “c”) : rotor 10 in Figure 1 of Barnett patent; DX-41; DX-1, page 3, 4;
locating means (element “d”) (for locating said movable means in a certain location with respect to said stator means) : light beams 26, photo cells 30, and control unit 18 in Figure 1 of DX-41, DX-1, page 5;
*661Element in Claim 7 Corresponding Structure in Barnett Specification
first energy means (element “e”) (responsive to movement of said movable means away from said certain location, for energizing said locating means) electromagnets 22 in Figure 1 of DX-41; DX-1, page 5;
second energy means : (element "f”) (for imparting rotation to said movable means; the said movable means being rotatable in response to energy from a second energy means) light energy 36, from light sources 34 on stator. Some light energy 36 is used to power rotor photocells 46. See Figure 1, DX-41; DX-1, page 6;
third energy means : (element “g”) (carried by the movable means; the said third energy means being actuated by said second energy means) photocells 46 on rotor vanes 38 send current to light bulb 40 in Figure 1, DX-41, DX-1, page 6;
indicating means : (element “h”) (carried by the stator means, responsive to energy of said third energy means, to indicate relative movement of the axis of the movable means with respect to the stator means) photocells 44 on stat- or with light beam 43 in Figure 1, DX-41; DX-1, page 7.
Once having analyzed the scope of “means” claim 7, by reference to the relevant portions of the Barnett specifications, claim 7 then must be read on the accused device. As explicated by Chief Judge Mar-key: “Determination of patent infringement requires two steps: the meaning of the claims must be learned from a study of all relevant patent documents; and the claims must be applied to the accused structures.” Caterpillar Tractor Co. v. Berco, S.P.A., 714 F.2d 1110, 219 U.S.P.Q. 185, 187 (Fed.Cir.1983).
Moreover, unforeseen improvements are still covered by the claims of a patent, despite the fact that they, by their very nature, have not been disclosed:
An applicant for patent is required to disclose the best mode then known to him for practicing his invention. 35 U.S.C. § 112. He is not required to predict all future developments which enable the practice of his invention in substantially the same way. [Caterpillar, [714 F.2d 1110] 219 U.S.P.Q. at 188.]
When analyzing claim 7, pursuant to the procedure indicated in Caterpillar, it is clear that elements “d”, “e”, “f” and “g”, supra, do not read on the accused Rockwell devices. In the disclosed patented device (the disclosed Barnett gyroscope), two distinct subsystems are required for elements “d” and “e”, i.e., the “locating means ” consisting of light beams 26, photocells 30 and control unit 18; and the “first energy means” consisting of electromagnets 22. In contrast, however, in the accused Rockwell gyroscopes, the functions of both the “locating means” (element “d”), and the “first energy means ” (element “e”), are performed by a single element, i.e., the stator octants. Therefore, elements “d” and “e” of claim 7 do not read on the accused Rockwell devices.
In regard to element “f”, the “second energy means ”, the disclosed patented device and the accused Rockwell ESG device again significantly differ. In the disclosed patented device, a single subsystem, i.e., the light energy 36 that falls on the rotor vanes 38, fulfills the requirement of the “second energy means ", which is element “f”. In contrast, however, in the accused Rockwell device, two separate subsystems, i.e., the electromagnetic coils, and the suspension servo, fulfill the functional requirement of the “second energy means”. Therefore, element “f” of claim 7 does not read on the accused Rockwell gyroscopes.
In regard to element “g”, the “third energy means carried by movable means ”, the disclosed patented device and the accused Rockwell ESG again significantly differ. In the disclosed patented device (the Barnett gyroscope), lightbulb 40 on rotor 38 fulfills the requirement of *662“third energy means carried by movable means”. In contrast, however, in the accused Rockwell ESG, no one separate element performs the sole function of the “third energy means carried by the movable means”. Plaintiff asserts, however, that the accused Rockwell ESG’s rotor, which has already been designated as fulfilling the movable means, also fulfills the third energy means. Plaintiff thus advances the novel theory that the accused Rockwell ESG’s rotor, performs the additional dual role fulfilling the function of the third energy means also. However, assuming, arguendo, that plaintiff is correct, the disclosed patented device and the accused Rockwell device still significantly differ because the accused device does not have a separate subsystem for the movable means and the third energy means, while, in contrast, the disclosed patented device contains separate subsystems for the movable means and third energy means. Thus, element “g” does not read on the accused Rockwell device.
Moreover, while the patentee is his own lexicographer, there is no rationale for using the claim language “energy means carried by the movable means” unless it is intended to refer to a specific active energy source, as opposed to the “rotor” itself, which is already covered by the “movable means”. The accused Rockwell ESG’s rotor, rather than being an active rotor as asserted by plaintiff, is more properly described as an inactive rotor, or a passive rotor. It is in fact merely a passive element which possesses energy which is impressed upon it. Finally, plaintiff’s attorney admitted indirectly that its literal infringement case was frail, at best. Plaintiff admits that “[t]he structure of the Rockwell ESG, in some respects, is different from that of the Barnett gyroscope disclosed in the illustrative embodiment.” See plaintiff’s post-trial brief at 14. Additionally, plaintiff’s substantial efforts have been focused upon establishing his infringement argument based on the doctrine of equivalents, rather than on literal infringement. See, e.g., plaintiff’s post-trial brief at 14-21.
Finally, since plaintiff did not assert, at trial, that Rockwell’s differences were but mere “unforeseen improvements”, we do not find that these significant differences were mere “unforeseen improvements”.
Doctrine of Equivalents
Since this court, having read claim 7 on the accused device, has not found literal infringement of the sole claim at suit, it must address the issue of whether or not infringement exists pursuant to the doctrine of equivalents.
The seminal genesis of the doctrine of equivalents, of course, was the historic 1853 case of Winans v. Denmead, 56 U.S. (15 How.) 329, 14 L.Ed. 330 (1853), where the Supreme Court first expanded the literal wording of a claim to capture an infringer. This was followed by a series of Supreme Court cases, ending with the much quoted 1950 case, Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 607-608, 70 S.Ct. 854, 855-56, 94 L.Ed. 1097, 85 U.S.P.Q. 328, 330-331 (1950), wherein the Supreme Court stated:
In determining whether an accused device or composition infringes a valid patent, resort must be had in the first instance to the words of the claim. If accused matter falls clearly within the claim, infringement is made out and that is the end of it.
But courts have also recognized that to permit imitation of a patented invention which does not copy every literal detail would be to convert the protection of the patent grant into a hollow and useless thing. Such a limitation would leave room for — indeed encourage — the unscrupulous copyist to make unimportant and insubstantial changes and substitutions in the patent which, though adding nothing, would be enough to take the copied matter outside the claim, and hence outside the reach of law. One who seeks to pirate an invention, like one who seeks to pirate a copyrighted book or play, may be expected to introduce minor variations to conceal and shelter the piracy. Outright and forth*663right duplication is a dull and very rare type of infringement. To prohibit no other would place the inventor at the mercy of verbalism and would be subordinating substance to form. It would deprive him of the benefit of his invention and would foster concealment rather than disclosure of inventions, which is one of the primary purposes of the patent system.
The doctrine of equivalents evolved in response to this experience. The essence of the doctrine is that one may not practice a fraud on a patent. Originating almost a century ago in the case of Winans v. Denmead, 15 How. 330, 14 L.Ed. 330, it has been consistently applied by this Court and the lower federal courts, and continues today ready and available for utilization when the proper circumstances for its application arise. “To temper unsparing logic and prevent an infringer from stealing the benefit of an invention” a patentee may invoke this doctrine to proceed against the producer of a device if it performs substantially the same function in substantially the same way to obtain the same result.” Sanitary Refrigerator Co. v. Winters, 280 U.S. 30, 42, 50 S.Ct. 9, 13, 74 L.Ed. 147. The theory on which it is founded is that “if two devices do the same work in substantially the same way, and accomplish substantially the same result, they are the same, even though they differ in name, form, or shape.” Machine Co. v. Murphy, 97 U.S. (7 Otto) 120, 125, 24 L.Ed. 935. The doctrine operates not only in favor of the patentee of a pioneer or primary invention, but also for the patentee of a secondary invention consisting of a combination of old ingredients which produce new and useful results, Imhaeuser v. Buerk, 101 U.S. (11 Otto) 647, 655, 25 L.Ed. 945, although the area of equivalence may vary under the circumstances. See Continental Paper Bag Co. v. Eastern Paper Bag Co., 210 U.S. 405, 414-415, 28 S.Ct. 745, 749, 52 L.Ed. 1122, and cases cited; Seymour v. Osborne, 11 Wall. (78 U.S.) 516, 556, 20 L.Ed. 33; Gould v. Rees, 15 Wall. 187, 192, 21 L.Ed. 39. The wholesome realism of this doctrine is not always applied in favor of a patentee but is sometimes used against him. Thus, where a 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, but nevertheless falls within the literal words of the claim, the doctrine of equivalents may be used to restrict the claim and defeat the patentee’s action for infringement. [Westinghouse v. Boyden Power Brake Co., 170 U.S. 537, 568, 18 S.Ct. 707, 722, 42 L.Ed. 1136. (Emphasis supplied).]
Plaintiffs assert that claim 7 of the Barnett patent reads on the accused Rockwell gyroscopes under the doctrine of equivalents. Defendant adversatively maintains that claim 7 is precluded by prosecution history estoppel, from encompassing the Rockwell gyroscopes through the doctrine of equivalents.
Thus, once a determination is made that the doctrine of equivalents is applicable, as this court has done, we must look to the prosecution history to determine limitations, if any, on its applicability, while at the same time assuming that the proposed construction of the claims is not inconsistent with the patentable scope to be accorded the claim in light of the prior art. Carman Industries, Inc. v. Wahl, 724 F.2d 932, 220 U.S.P.Q. 481 (Fed.Cir.1983).
The Federal Circuit, however, has clearly rejected the proposition that virtually any amendment of the claims at issue will create an estoppel limiting the scope of protection accorded to those claims. The doctrine of equivalents, therefore, should not be limited solely to those few claims allowed exactly as originally filed. Hughes Aircraft Co. v. United States, 717 F.2d 1351, 219 U.S.P.Q. 473 (Fed.Cir.1983). A priori, whenever the doctrine of prosecution history estoppel is invoked, the court must scrutinize not only what was surrendered, but also the underlying reasons for such surrender. The scope of the estoppel must bear a direct relationship to the extent to which the prior art was avoided by the amendment or argument in question. Bayer A. G. v. Duphar *664Int’l Research B.V., 738 F.2d 1237, 221 U.S.P.Q. 1056 (Fed.Cir.1984); Fromson v. Advance Offset Plate, Inc., 720 F.2d 1565, 219 U.S.P.Q. 1137 (Fed.Cir.1983); ACS Hospital Systems v. Montefiore Hospital, 732 F.2d 1572, 221 U.S.P.Q. 929 (Fed.Cir.1984). For instance, an amendment needed solely to overcome a rejection under 35 U.S.C. § 112, will not ordinarily create an estoppel. Caterpillar Tractor Co. v. Becco, S.P.A., etc., 714 F.2d 1110, 219 U.S.P.Q. 185 (Fed.Cir.1983). In response to those who would take a mechanistic view of claim interpretation under such an extreme interpretation of the doctrine of file wrapper estoppel, now more aptly named prosecution history estoppel, see Hughes Aircraft Co. v. United States, 717 F.2d 1351, 219 U.S.P.Q. 473, 481 (Fed.Cir.1983), wherein Chief Judge Markey pointed out that:
We ... reject that [narrow] view as a wooden application of estoppel, negating entirely the doctrine of equivalents and limiting determination of the infringement issue to consideration of literal infringement alone. [Hughes, 717 F.2d 1351, 219 U.S.P.Q. at 481. (Emphasis supplied).]
The realistic requirement for a case-by-case evaluation of the circumstances surrounding an amendment was further explicated in Hughes Aircraft:
Amendment of claims is a common practice in prosecution of patent applications. No reason or warrant exists for limiting application of the doctrine of equivalents to those comparatively few claims allowed exactly as originally filed and never amended. Amendments may be of different types and may serve different functions. Depending on the nature and purpose of an amendment, it may have a limiting effect within a spectrum ranging from great to small to zero. The effect may or may not be fatal to application of a range of equivalents broad enough to encompass a particular accused product. It is not fatal to application of the doctrine itself. [Hughes Aircraft Co. v. United States, 717 F.2d 1351, 219 U.S.P.Q. 473, 481 (Fed.Cir.1983).]
Thus, a careful analysis must be made of the prosecution history of the Barnett patent to determine if its prosecution history preclude plaintiffs, subsequently, from utilizing the doctrine of equivalents in an infringement suit to extend the scope of his claim. As the court in Coleco Industries stated:
A response to an examiner’s office action may include an amendment accompanied by remarks or it may include only remarks. Because applicant’s ultimate goal in submitting amendments and offering arguments in support thereof is the securing of a patent, we find no reason not to extend the traditional es-toppel doctrine beyond estoppel by amendment to estoppel by admission. Therefore, whenever a patentee utilizes the doctrine of equivalents in an infringement suit to extend the scope of this claims, he opens his case to rebuttal based on any statement he made on the record during prosecution. [Coleco Industries, 573 F.2d 1247, 1258, 65 CCPA 105 (1978).]
As stated in Coleco, prosecution history estoppel establishes a limitation on the doctrine of equivalents and may arise from any statement made by the applicant during the prosecution of the patent application.
In the ease at bar, arguments were asserted by plaintiffs that clearly restricted plaintiffs’ interpretation of its own claims. Plaintiffs’ patent application was filed on April 5,1963, by Eugene R. Barnett and his father Willard L. Barnett, the latter since deceased. In the first office action, dated April 7, 1964, the Patent Office cited four United States patents3 and rejected all eleven claims in the Barnett application. These four patents disclosed various magnetic and photoelectric means in gyroscope devices. In an amendment filed October 7, 1964, plaintiffs argued that the references *665did not disclose the use of a magnetic means to locate the gyroscope member as the Examiner had asserted, and that the references also failed to disclose several specific features of the plaintiffs’ gyroscope. Of particular relevance to issued claim 7, applicants stated that these references failed to disclose the use of an energy conversion means located on the rotor which converts some of the light used to spin the rotor into electrical energy to power the lamp, which projects the spin-axis light ray from the rotor to the stator.
In response to plaintiffs’ arguments, the Examiner issued a second office action dated December 28, 1966, in which ten additional prior art patents were cited to reject all 10 claims, as obvious under 35 U.S.C. § 103. The principal reference was Parker, U.S. Patent No., 2,919,583, which also taught a magnetically supported gyroscope. The Parker gyroscope is described in Figure 1 of the patent, reproduced supra.
The Examiner, in explicating the relevancy of Parker, explained how Parker interacted with Beeh, Simon, and Elwell, to render the claimed invention as a whole, obvious to one skilled in the art:
Parker shows a magnetically supported gyroscope and a photoelectric system for sensing and correcting the position of the rotor. Parker provides electron beam means aimed tangentially of the rotor for driving the rotor without mechanical contact. Beeh shows a radiometer drive means for rotating a light modulator which is analogous to Parker’s drive means. Simon and El-well et al show photoelectric means for sensing and correcting the axis of rotation of the rotor which are the equivalents of applicants’ means. The use of photovoltaic cells to generate power without mechanical connections to a generator is too well known to require additional support, but it is noted that MarH-son shows this concept. Thus, it is held that the prior art renders obvious the instant invention as defined by the presented claims. [Emphasis supplied.]
More specifically, the Parker gyroscope contains a spherical rotor IB which is mounted in an enclosure 25, and is suspended by an electromagnetic field created by electromagnets 11, 12, 13 and 14. The translational position of the rotor 10 is maintained by a photoelectric control system comprised of lamp 18 and photoelectric cell 22, in essentially the same manner as the Barnett gyroscope. The rotor is driven by an electron gun 27 which projects a beam of electrons 28 tangentially on the surface of the rotor. The position of the spin axis is determined by photoelectric detector 71 which detects the modulation pattern of the beam of light projected by lamp IB upon the rotor and reflected by a band having different optical properties placed on the surface of the rotor.
Plaintiffs responded to the December 28, 1966, Office Action by filing an amendment dated March 27, 1967. In this amendment, the plaintiffs explicitly distinguished their invention from the gyroscope shown in Parker, by focusing the Examiner’s attention to their rotor-position pickoff scheme. The following statements were made by plaintiffs:
A. {Double-function of rays 36)” In applicants’ device, the same energy beams (36) provide the double function of both imparting rotation of the rotor and energizing the rotor circuit (48) which then energizes energy-means (40) which projects energy ray (43) for the rotation sensing cells (44). In contrast, the primary reference Parker uses an electron beam 28 to impart rotation to his rotor 10, but does not use that electron beam for the other purpose stated above as in applicants’ invention; and the Simon and Elwell patents, cited by the Examiner as sensing rotation-wandering, do not utilize those rotation-sensing-rays in any relationship whatever to the rotation-imparting rays, and certainly not in the specific relationship of the rotation-sensing-rays being actually energized by the rotation imparting rays.
This distinction, referred to sometimes as Distinction “A ” hereinafter, is a fea*666ture of patentable novelty present in most of the claims, as will be noted.
B. {Radiometer Actuation): While not here asserting that radiometer-actuation in and of itself would render a claim patentable, here the radiometer-actuation is a part of a combination concept pointed out in “A” above, that is, the radiometer actuation is a combination in which the energy rays for the radiometer-actuation also energize the rotor circuitry and the rotor energy means which project the rotation sensing beam 43. [Emphasis supplied.]
On June 23, 1967, the Examiner allowed claims 1-3, 5-10, and 12 “in view of applicants’ persuasive and clearly presented arguments.” [Emphasis supplied.] Application claims 7-10 and 12 became claims 6-10 in the patent in suit which issued on April 23, 1968. Application claim 8 became claim 7 in suit.
Thus, in describing the exact nature of their invention, applicants focused the attention of the Examiner on the “double-function of rays 36 ”, i.e., energy beams 36, used to develop the rotor-position pick-off information in the Barnett gyroscope. As discussed below, there is no double-functioning or even single-functioning energy beams, used for rotor-position pickoff information or any other purpose, in the Rockwell gyroscope.
Accordingly, the applicants are precluded by the doctrine of prosecution history estoppel from attempting to enlarge the scope of the claims, in particular claim 7 in suit, to encompass a gyroscope such as the Rockwell ESG, which utilizes a capacitive bridge, passive pickoff scheme, and not the double-functioning energy beams and active energy conversion rotor-position pickoff scheme, which they stressed to the Patent Office, as their “feature of patentable novelty present in most of the claims.” As clearly seen, this “feature” was also embraced in issued claim 7, the sole claim in issue.
A Paper Patent in a Crowded Art
There is an additional reason why claim 7 cannot be expanded, under the doctrine of equivalents, to read on the Rockwell gyroscope. The gyroscope which is disclosed in the Barnett patent has never been completely constructed by its inventors or anyone else. Consequently, the patent in suit is a “paper patent” and is entitled to a very narrow range of equivalents. Caldwell et al. v. United States, 202 Ct.Cl. 423, 432, 481 F.2d 898, 903, 175 U.S.P.Q. 44, 48 (1973). Patentees are not entitled to preempt the entire field of electrostatically supported gyroscopes simply because they may have conceived a peculiar photoelectric rotor-position pickoff concept intended to be utilized in such a gyroscope. “The underlying basis for the application of the paper patent theory as applied to infringement, is that an inventor is not entitled to restrain the progress of his art by failing to use his invention.” Glendenning v. Mack, 159 F.Supp. 665, 668, 116 U.S.P.Q. 249, 252 (D.Minn.1958). Moreover, it is clear that the design and manufacture of a high precision gyroscope had become a very crowded art by April 1963, when the Barnett application was filed, and the Barnett light beam gyroscope only added a specific optical photoelectric pickoff concept which represented a small addition to this art.
More specifically, plaintiffs did advance the limited relevant art in the field of gyroscopes, i.e., developing an active free rotor. While it is recognized that “active ” spherical free rotors do not constitute a recognized field of art, nevertheless, only the Barnett gyroscope appears to have an active spherical free rotor. Knowledge of the orientation of the rotor’s axis vis-a-vis the stator is absolutely essential for navigation. Various techniques have been proposed, tried, and used over the years to obtain this information, often referred to as the determination of “angular pickoff” or to determine a “rotor’s spin-axis position ”, accurately.4
*667Where the rotor plays merely a “passive ” role in the process, e.g., by reflecting light from its distinctively marked surface to observation points on the stator, the rotor does not expend any of its own energy nor convert any of its own energy to another form of energy, which in turn is used to provide axial position information. While the plaintiffs argue that some of the capacitance energy, in the Rockwell ESG, is converted to another form of energy, namely “heat”, this “heat energy” is not involved in delivering rotor-position pickoff information to the stator. Therefore, the accused Rockwell ESGs are more truly defined as “passive” spherical rotors, rather than “active” spherical rotors. Other examples of gyroscopes using passive rotors, are those taught by Kunz and Parker.
On the other hand, where the gyroscope has an “active” rotor, i.e., one that must receive, convert, and expend energy to transmit its axial position information to the stator, it becomes necessary to continuously replenish this energy of the rotor from the stator for prolonged accurate performance of the gyroscope. Since no evidence was adduced at trial that the patented Barnett gyroscope was not the first to employ an “active’’spherical free rotor in a gyroscope, we assume for the purposes of this trial, that the patented Barnett gyroscope was the first as asserted by plaintiff.
Therefore, the Barnett invention, arguably, can be asserted to be a “pioneer” invention, at least as it applies to the field of gyroscopes containing active spherical free rotors. Its status as a “paper patent”, however, has an equal and opposite countervailing effect on its right to assert that a broader than normal doctrine of equivalents should be accorded it. As such, left solely with the effects of plaintiffs' prosecution history estoppel, we conclude that plaintiffs may not utilize the doctrine of equivalents to expand the literal meaning of issued claim 7, as defined by the Barnett specification itself, in order to read on the accused Rockwell gyroscopes that do not include an energy source on its rotor, and that do not utilize the “double-function of [its energy] rays”, which were pointed out as “Distinction A” in the prosecution history of the Barnett patent.
Even assuming, arguendo, that the doctrine of equivalents is applicable, the elements of the Rockwell ESG are not equivalent or interchangeable with the elements of claim 7 of the Barnett patent. As the Federal Circuit explicated in Thomas & Betts Corp. v. Litton Systems, Inc., 720 F.2d 1572, 1579, 220 U.S.P.Q. 1, 6 (Fed.Cir.1983):
[T]he test of equivalency extends beyond what is literally stated in defendant patentee’s specification to be equivalent and encompasses any element which one of ordinary skill in the art would perceive as interchangeable with the claimed element.
While this court finds that the doctrine of prosecution history estoppel clearly precludes plaintiffs from availing themselves of it, we, nevertheless, address the doctrine of equivalents issue because the resolution of the complete infringement issue may also overlap upon the resolution of other issues. In this same context, Chief Judge Markey stated in Lindemann Maschinenfabrik GMBH v. American Hoist and Derrick Co., 730 F.2d 1452, 1463-64, 221 U.S.P.Q. 481, 489-90 (Fed.Cir.1984):
A district court should decide validity and infringement and should enter a judgment-on both issues when both are raised in the same proceeding. Stratoflex v. Aeroquip, 713 F.2d 1530, 218 U.S.P.Q. 871 (Fed.Cir.1983). To enter judgment on less than all disposi-tive issues can be inefficient, risking as it does the necessity of the district court and the parties undertaking participation *668in another long and costly proceeding. [Emphasis supplied.]
A. The Rotor-Position Pickoff Concepts Are Entirely Different
As previously noted, the well-known test of equivalents in patent law requires that the corresponding means in the accused and patented devices be: (1) essentially the same, (2) accomplish substantially the same result, and (3) do so in substantially the same way. Coleco Industries, Inc. v. U.S. Int’l Trade Comm’n, 573 F.2d 1247, 1254, 65 CCPA 105 (1978), Graver Tank and Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 608, 70 S.Ct. 854, 856, 94 L.Ed. 1097 (1950).
In order to develop rotor-position pickoff information in the Barnett gyroscope, the rotor 10 carries a photoelectric active energy- conversion system in which a portion of the light energy, from light energy beams 36, utilized to drive the rotor through the radiometer effect, is converted into electrical current by photocells mounted on the rotor. This current, in turn, is used to energize a lamp mounted on a spin axis of the rotor 10 to project a narrow light beam 43 along the spin axis toward the stator. The spin axis light beam 43 impinges upon an array of photocells 44 mounted on the stator, and it is the particular position at which the light beam impinges upon the photocell array, that provides an indication of the rotor orientation.
In the Rockwell gyroscope, on the other hand, there is no active energy conversion apparatus mounted on its rotor for rotor-position pickoff or any other purpose. The effect of the small wobble in the geometric envelope of its rotor as it spins, is to redistribute the energy stored in the capacitive electrostatic suspension field between the rotor and stator, and it is this redistribution which is sensed as a modulation signalled by the suspension electrodes, and which is interpreted to develop the rotor-position pickoff information. Moreover, the small amount of capacitance energy that is “converted” into heat energy (during redistribution of electrical charges on the rotor), is not utilized in any manner to provide information regarding the rotor spin-axis position in the Rockwell ESG pickoff scheme. The “heat energy” merely dissipates and plays no role whatsoever in the Rockwell ESG rotor spin-axis determination. Whereas, the converted “electrical current” (in the Barnett gyroscope) is of cardinal importance in the Barnett rotor spin-axis position pickoff scheme.
Therefore, the following important differences exist between the Rockwell rotor-position pickoff concept and the Barnett rotor-position pickoff concept:
(1) In the Barnett gyroscope, the rotor-position pickoff information is developed through the receipt and conversion by the rotor of light energy into electrical current and back into light energy, which is projected from the rotor to the stator, by a special apparatus mounted on the rotor. There is no similar conversion of energy by the rotor in the Rockwell gyroscope. The effect of the rotor rotation is to mechanically alter the distribution of the electrical energy capacitively stored within the electrostatic suspension field, and it is this redistribution from which the pickoff information is derived. Moreover, the side effect of “heat”, which plaintiff labels as a “conversion”, is ipse dixit, and, this “heat” does not play even a scintilla of a role in providing rotor spin-axis position information in the Rockwell ESG.
(2) The pickoff information is derived from the capacitive electrostatic suspension field in the Rockwell gyroscope. In the Barnett gyroscope, on the other hand, there is no interaction between the electromagnetic suspension field and the light beams and apparatus which provide the pickoff information.
Thus, while the pertinent function performed in the Rockwell and Barnett gyroscopes is essentially the same, namely, development of the rotor-position pickoff information, this function is inherent in any gyroscope, and is performed in the Rockwell gyroscope with entirely different structures in a wholly different manner. *669Consequently, it must be concluded that the rotor-position pickoff structure of the Rockwell gyroscope is not equivalent to that of the Barnett gyroscope and claim 7 of the patent is therefore not infringed.
B. The Rockwell ESG Gyroscope Uses Neither The Multi-Element Structure Of The Barnett Gyroscope Nor An Equivalent of Several Of The Elements
In claim 7, each of the major functions of the Barnett gyroscope is claimed as a separate and distinct element. For example, the magnetic suspension field is denominated as a “locating means” and the photoelectric position control system is the “first energy means”. The radiometer drive system is identified as the “second energy means”. The rotor-position pickoff system is comprised of a “third energy means” on the rotor which interacts with the “indicating means” on the stator. Conversely, all of these functions are performed in the Rockwell gyroscope by the single element corresponding to the electrostatic suspension field. Additionally, there are no elements in the Rockwell gyroscope which are separate from the “locating means”, and which correspond to the first, second, and third “energy means” recited in claim 7.
The multi-element nature of the Barnett gyroscope is apparent from the fact that it cannot be used as a free rotor or “strap down” gyroscope, whereas the Rockwell gyroscope has been used successfully as a strap down gyroscope in the experimental MICRON (also known as the N-73 and AN/ASN-122) aircraft gyroscope. A free rotor gyroscope is characterized by the ability of the rotor to assume and maintain any angular relationship with respect to the stator. Therefore, the stator of such a gyroscope can be “strapped down”, i.e., hard mounted to the vehicle in which it is operating. Conversely, the angular relationship between the rotor and stator of the Barnett gyroscope must be maintained within certain narrow limits. This limitation is due to the fact that the interior surface of the Barnett stator must carry several elements at mutually exclusive locations around its circumference. These elements are: (1) lights 28 and photocells 30 of the translational position (suspension) control system; (2) lamp array 34 which produces the radiometer drive beam; and (3) photocell array 44 which detects the position of rotor spin axis light beam 43.
While the Barnett patent discloses the use of light beams of different frequencies to keep the separate functions from interfering with one another, it contains no teaching of how the gyroscope could function when the position of the rotor causes the spin axis light beam to fall on lights 28, photocells 30, or lamp array 34 of the position control and rotor drive systems. In other words, to operate successfully, the stator of the Barnett gyroscope must be controlled so that it moves with the spin axis of the rotor to keep the spin axis light beam focused upon the limited area of the pickoff photocell array 44. Consequently, the Barnett gyroscope cannot, without undue experimentation, be used as a free rotor or strap down gyroscope as can the Rockwell gyroscope.
The Rockwell gyroscope also does not utilize an equivalent of each missing element. The Barnett gyroscope uses an electromagnetic field to suspend its rotor, while the Rockwell gyroscope uses an electrostatic field for this purpose. There is no teaching in the specification and drawings of the Barnett patent of the use of electrostatic, rather than electromagnetic suspension. In fact, the large rotor to stator gap shown in the patent is not compatible with the use of electrostatic suspension. Moreover, the side effects associated with the two types of fields are entirely different. Consequently, it must be concluded that the electrostatic suspension concept used in the Rockwell gyroscope operates in an entirely different manner, and is not the equivalent of the electromagnetic suspension concept used in the Barnett gyroscope.
In addition, the Barnett gyroscope uses a photoelectric scheme to sense the rotor’s translational position, and send corrective information to the suspension electromag*670nets. The Rockwell gyroscope, on the other hand, uses the capacitive bridge effect of the electrostatic suspension field for this purpose. These two principles of operation are very different and are not equivalent.
Finally, the Barnett gyroscope uses the radiometer effect to drive its rotor, and has no provision for speed control, whereas the Rockwell gyroscope is furnished with electromagnetic induction coils to bring the rotor to its operational speed at startup, and depends upon the small rotational moment provided by the electrostatic suspension field interacting with the slight mass unbalance feature of the rotor, to maintain the proper speed during normal operation. Again, these principles of operation are very different in the two gyroscopes and cannot validly be considered as equivalent.
The fundamental differences in the structure and mode of operation of the rotor suspension, rotor, drive, and rotor-position pickoff concepts utilized in the Barnett and Rockwell gyroscopes were established by the testimony at trial of the plaintiffs’ expert witness as well as that of defendant’s expert witness. There was no substantial disagreement between their testimony and opinions in this regard. There was no evidence to the contrary. Accordingly, the clear and convincing weight of the evidence is, and it must be concluded, that the Rockwell electrostatically supported gyroscopes are not equivalent to the Barnett light beam gyroscope. Accordingly, claim 7 of the patent in suit is not infringed.
C. Plaintiffs’ Decision Not To Call The Plaintiff-Patentee As A Witness Also Gives Rise To The Inference That Rockwell’s ESG Gyroscopes Are Not Equivalent To The Barnett Gyroscope
Plaintiff Eugene R. Barnett is the surviving named inventor. From an early date, plaintiffs had indicated in their pretrial papers that Mr. Barnett would be called as one of their witnesses at trial. However, only one week prior to trial, plaintiffs changed their minds and indicated that he would not be called as a witness. Moreover, he was not in attendance in Los An-geles during the trial, therefore discouraging defendant from even calling him as a hostile or rebuttal witness.
The testimony of Mr. Barnett as to how he and his father developed the gyroscope concepts reflected in the patent in suit, would have been illuminating and helpful toward understanding the true nature and scope of their invention. Accordingly, defendant sought to introduce into evidence the deposition testimony of Mr. Barnett in this regard, but was strenuously opposed by plaintiffs’ counsel who asserted that what the inventors actually had in mind, would be irrelevant and, could prejudice the court’s decision of the case:
Mr. Maxwell: And what I am saying here is that, just leafing through it now, I think what [defendant’s attorney] is talking about is the development of the device itself. And I think that the statement that “what could be more relevant?”, I think is superficially pleasant, but I think it’s inaccurate. I think the court has no real interest really in what he had in mind. What the court will see is the patent and the device. We are not seeking to go beyond the application date; so the shop notes and what this inventor was doing prior to the date of the application really is, as we see, of no concern, to the court or to the defendant.
* * * * s¡e
Mr. Daigle: Your Honor, there are at least a dozen requested findings on both sides that are directly in conflict. It is primarily the parties’ two conflicting views of what type of rotor position pick-off system is used. And I think to enlighten what the inventor envisioned his pickoff system to be, his words are invaluable and of guidance to the court.
We do, of course, have what is disclosed in the patent. But I think how the inventor went about coming up with those ideas gives an idea of its basic nature.
We would have no problem or a necessity of going beyond this if this case involved solely the embodiment in the patent in suit. But, obviously, the ac*671cused device is not that. And they are telling [us] to expand it somewhat beyond what is in the patent. And I think as evidence of what the invention truly was, the inventor’s own words and description are of value, even if they are not being used to take the invention date back beyond the filing date.
Mr. Maxwell: We think that it is possible that they could be persuasive. But we think if they were persuasive, they would be improperly persuasive. We think they are not of value.
We think that what is of value is the patent and any reasonable equivalence of the patent as the evidence adduced to the court shows are reasonable equivalents.
What the inventor thought he was doing may be grandiose. It may be less than he did. And the cases are quite clear that he could have created something more than he thought he was creating or something less than he was creating.
You know, what is becoming apparent to me is that it may be that this testimony, which I would have to see again, could conceivably prejudice the court in that fashion. I really think it’s clear that there isn’t any relevance in this testimony. The claims are very broad claims; and we seek to establish those claims, and the embodiments, and the reasonable equivalence thereof through the testimony as the plaintiff sees fit to have that testimony adduced. Of course, the defendant has the opportunity to present its own case. But the testimony of what the inventor thought he was doing, I don’t see how really it can be helpful to the court. (Tr. 44-46)
It has long been a well settled principle of evidence that, where a party fails to call a witness available to him who has knowledge of material facts, the court may draw the inference that the testimony of the witness concerning those facts would have been unfavorable to the party. Culbertson v. The Steamer Southern Belle, 59 U.S. (18 How.) 584, 588, 15 L.Ed. 493 (1855). This is especially applicable where the witness is a party. “The rule is that: ‘The nonappearance of a litigant at the trial or his failure to testify as to facts material to his case as to which he has especially full knowledge creates an inference that he refrained from appearing or testifying because the truth, if made to appear, would not aid his contention.’ ” United States v. Fields, 102 F.2d 535, 537-538 (8th Cir.1939); see Barnett v. United States, 319 F.2d 340, 344 (8th Cir.1963). As has been discussed, plaintiffs argued to the Patent Office that their invention resided solely in the photoelectric pickoff concept in which light beams are used to drive the rotor and power a separate active electrical circuit mounted on the rotor which converts the light into current which energizes a lamp to send out a light ray along the rotor spin axis. The failure of plaintiff Eugene R. Barnett to appear at trial to explain the true scope and nature of his invention gives rise to the inference that this was all there was to it and that the Rockwell electrostatic gyroscope is not an equivalent.
D. Claim 7 Would Read On The Parker Prior Art Gyroscope And Be Invalid If Not Limited To The Light Beam Energy Conversion Pickoff Concept
In an Office Action dated December 28, 1966, the Examiner cited Parker, U.S. Patent No. 2,919,583, which discloses a magnetically supported gyroscope that is very similar to the Barnett gyroscope. A comparison of the elements of claim 7, with the corresponding structures disclosed in the Barnett patent and the Parker patent, is set forth hereinbelow:
Claim 7 Barnett Gyroscope Parker Gyroscope
(a) A control means comprising: gyroscope
(b) a stator means stator 12 in casing 14 enclosure 25
(c) and a movable means; rotor 10 rotor 10
(d) locating means for locating said movable means in a certain location with respect to said stator means; electromagnetic electromagnetic field produced by field produced by electromagnets 22 electromagnets acting on rotor 10 11-14 acting on rotor 10
*672Claim 7 Barnett Parker Gyroscope Gyroscope
(e) first energy means, responsive to movement of said movable means away from said certain location, for energizing said locating means; photoelectric system in which lights 28 and photocells 30 detect translational position of rotor 10 photoelectric system in which light source 16 and photoelectric detector 17 detect translational position of rotor 10
(f) second energy means for imparting rotation to said movable means; the said movable means being rotatable in response to energy from a second energy means; light beam system in which lights 34 on stator 12 transmit energy to vanes 38 on rotor 10 to cause rotation through the ratio-meter effect electron beam system in which electron gun 27 transmits energy tangentially on rotor 10 to spin up the rotor initially
(g) third energy means carried by the movable means; the said third energy means being actuated by said second energy means photoelectric system in which light 42 is powered by the energy conversion system carried by the rotor to project the focused, narrow light beam 43 along the spin axis of the rotor photoelectric system in which light is reflected and modulated by the rotor which contains predetermined spots having optical qualities different from the general surface of the rotor and are detected as a band indicative of spin axis orientation
(h) indicating means carried by the stator means, responsive to energy means, to indicate relative movement of the axis of the movable means with respect to the stator means; photoelectric system in which the orientation of light beam 43 is detected by photocell array 44 mounted on the stator photoelectric system in which the light pattern reflected by the rotor is detected by photoelectric detector 71
In order to distinguish and preserve the subject matter of claim 7 from Parker, it is necessary to interpret the words “third energy means carried by the movable means” in element (g) as referring to a discrete energy device, such as the elee-trict circuit disclosed in the Barnett gyroscope, which is physically installed on and carried by the rotor, i.e., it must be something other than the rotational energy of the rotor. This is the interpretation which would be given to these words by one skilled in the art. Therefore, this comparison demonstrates that, if the claim in suit is broadened beyond the light beam energy conversion rotor-position pickoff concept, which Barnett argued to the Patent Office to overcome prior art, claim 7 would read on the prior art Parker disclosure, and be invalid.
A construction of the claims that would render the patent valid, however, is favored over a construction that would render the claims invalid. Roberts Dairy Co. v. United States, 208 Ct.Cl. 830, 869, 530 F.2d 1342, 1367 (1976). In order to avoid invalidity due to the readability of claim 7, if read too broadly, on the gyroscope disclosed in the Parker patent, claim 7 must be limited to the photoelectric active energy conversion rotor-position pickoff concept disclosed in the Barnett patent. Therefore, it must then be concluded that narrowed claim 7 is not infringed by the Rockwell gyroscope with its capacitive mass-unbalance-modulation rotor-position pickoff concept. Dominion Magnesium Ltd. v. United States, 162 Ct.Cl. 240, 252, 320 F.2d 388, 396, 138 U.S.P.Q. 306, 312 (1963); Int’l Glass Co. v. United States, 187 Ct.Cl. 376, 394-395, 408 F.2d 395, 405, 159 U.S.P.Q. 434, 442-443 (1969).
Validity
It is now clear beyond peradventure that: (1) the presumption of validity is never annihilated, destroyed, or even weakened, regardless of what facts are in the record, e.g., even with the introduction of prior act more pertinent than that considered by the examiner; (2) that it is clearly a statutory procedural device which assigns to the party asserting invalidity the burden of proving invalidity; and (3) that the burden of persuasion is, and remains always, on the party asserting invalidity. ACS Hospital Systems, Inc. v. Montefiore Hospital and Wells National Services Corp., 732 F.2d 1572, 1574-75, 221 U.S.P.Q. 929 (Fed.Cir.1984); Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 1534, 218 U.S.P.Q. 871, 875-876 (Fed.Cir.1983); Stevenson v. U.S. Int’l Trade Comm’n., 612 F.2d 546, 551, 67 CCPA 109, 204 U.S.P.Q. 276, 281 (1979); Solder Removal Co. v. U.S. Int’l Trade Comm’n., 582 F.2d 628, 632-633, 199 U.S.P.Q. 129, 132-133, 65 CCPA 120 (1978); Connell v. Sears, Roebuck & Co., 722 F.2d 1542, 220 U.S.P.Q. 193 (Fed.Cir.1983); Medtronic, Inc. v. Cardiac *673Pacemakers, Inc., 721 F.2d 1563, 220 U.S.P.Q. 97 (Fed.Cir.1983); Kalman v. Kimberly-Clark Corp., 713 F.2d 760, 773-774, 218 U.S.P.Q. 781, 790 (Fed.Cir.1983).
As eloquently explicated by Chief Judge Markey in Stratoflex, 713 F.2d at 1534:
The presumption, like all legal presumptions, is a procedural device, not substantive law. It does require the de-cisionmaker to employ a decisional approach that starts with acceptance of the patent claims as valid and that looks to the challenger for proof of the contrary. Thus, the party asserting invalidity not only has the procedural burden of proceeding first and establishing a prima-fa-cie case, but the burden of persuasion on the merits remains with that party until final decision. The party supporting validity has no initial burden to prove validity, having been given a procedural advantage requiring that he come forward only after a prima-facie case of invalidity has been made. With all the evidence in, the trial court must determine whether the party on which the statute imposes the burden of persuasion has carried that burden.
Introduction of more pertinent prior art than that considered by the examiner does not, heretofore, “weaken” or “destroy” the presumption. Nor does such introduction “shift” the basic burden of persuasion. The presumption continues its procedural, burden-assigning role throughout the trial. Such introduction can, of course, facilitate the validity challenger’s carrying of that burden. It would require one supporting validity to come forward with countervailing evidence, as would the introduction of any evidence tending to establish invalidity. In the end, the question of whether all the evidence establishes that the validity challenger so carried his burden as to have persuaded the decision-maker that the patent can no longer be accepted as valid. [(Footnotes omitted). (Emphasis supplied).]
The applicable statute, 35 U.S.C. § 282, in relevant part, reads as follows:
A patent shall be presumed valid. Each claim of a patent (whether in independent or dependent form) shall be presumed valid independently of the validity of other claims; dependent claims shall be presumed valid even though dependent upon an invalid claim. The burden of establishing invalidity of a patent or any claim thereof shall rest on the party asserting it. [Emphasis supplied.]
If that burden is not successfully carried by the party asserting invalidity, the trial court need only so state. It need not once more declare a patent “valid”, which was and still is valid, because the burden of proof of invalidity was not carried by the asserting party. Stratoflex, 713 F.2d at 1534 n. 3; 35 U.S.C. § 282.
In its post-trial brief and at trial, the government’s sole argument for invalidity was its assertion regarding plaintiffs’ noncompliance with the first paragraph of 35 U.S.C. § 112, i.e.,:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention. [Emphasis supplied.]
Thus, it is the government’s contention that claim 7 is invalid because the Barnett patent fails to disclose sufficient information to enable one skilled in the art to construct a workable and useful gyroscope. More specifically, while quoting only the first paragraph of section 112,5 defendant asserts that, without the assistance of post-filing-date state of the art, the Barnett gyroscope is unworkable. We disagree.6
*674The first paragraph of section 112 provides, in relevant part, that “the specification shall contain a written description of the invention”. 35 U.S.C. § 112 (1976). While not always apparent at first blush, this “description” requirement is actually separate and distinct from the “enablement” requirement. See In re Barker, 559 F.2d 588, 591, 194 U.S.P.Q. 470, 472 (CCPA 1977), cert. denied, 434 U.S. 1064, 98 S.Ct. 1238, 55 L.Ed.2d 764, 197 U.S.P.Q. 271 (1978).
In addressing the “written description” requirement of the first paragraph of section 112, the CCPA,7 in In re Edwards, stated:
The function of the description requirement is to ensure that the inventor had possession, as of the filing date of the application relied on, of the specific subject matter later claimed by him. To comply with the description requirement it is not necessary that the application describe the claimed invention in ipsis verbis; all that is required is that it reasonably convey to persons skilled in the art that, as of the filing date thereof, the inventor had possession of the subject matter later claimed by him. [In re Edwards, 568 F.2d 1349, 1351, 1352, 196 U.S.P.Q. 465, 467 (CCPA 1978) (Citations omitted) (Emphasis supplied).]
In determining whether the “written description” is adequate, resort may also be had to the use of knowledge possessed by those skilled in the art, at the time the application was filed, to supplement the disclosure. In re Lange, 644 F.2d 856, 209 U.S.P.Q. 288 (CCPA 1981). In delineating this proposition of law, the CCPA stated:
[T]he disclosure in question must be read in light of the knowledge possessed by those skilled in the art, and that knowledge can be established by affidavits of fact composed by an expert, and by reference to patents and publications available to the public prior to applicant’s filing date. [In re Lange, 644 F.2d at 863, 209 U.S.P.Q. at 294 (Citations omitted) (Emphasis supplied).]
Defendant, at trial, did not appear to assert, nor affirmatively carry its burden, that plaintiffs’ “written description” was specifically inadequate. The record, in fact, is substantially devoid of any specific attempt by defendant to address this specific burden. We therefore must conclude that defendant has not borne its burden, i.e., has not come forward with a prima facie case of invalidity due to noncompliance with the “written description” requirement of the first paragraph of § 112.
Rather, defendant’s main thrust regarding asserted invalidity of claim 7 is based on the contention that the enablement requirement of the first paragraph has not been satisfied. We disagree.
The relevant portion of § 112, first paragraph, reads:
The specification shall contain a written description ... of the manner and process of making and using [the invention], in such full, clear, concise, and exact terms as to enable any person skilled in the art ... to make and use the same____8
The question of enablement is perhaps the most significant inquiry under the first paragraph of § 112. As the CCPA stated in In re Moore:
What is of maximum concern in any analysis of whether a particular claim is supported by the disclosure in an application is whether that disclosure contains sufficient teaching regarding the subject matter of the claims as to enable one skilled in the pertinent art to make and to use the claimed invention. These two requirements, “how to make” and “how *675to use” have sometimes been referred to in combination as the “enablement” requirement ... The relevant inquiry may be summed up as being whether the scope of enablement provided to one of ordinary skill in the art by the disclosure is such as to be commensurate with the scope of protection sought by the claims.9
Moreover, as noted earlier, post-filing-date state of the art may not be used by plaintiffs to satisfy the “enablement” requirement. W.L. Gore & Associates, Inc., supra. Assuming, however, some experimentation were needed, a patent is still not invalid because of a need for experimentation. Minerals Separation, Ltd. v. Hyde, 242 U.S. 261, 270-271, 37 S.Ct. 82, 86, 61 L.Ed. 286 (1916). A patent is invalid only when those skilled in the art are required to engage in undue experimentation to practice the invention. In re Angstadt, 537 F.2d 498, 503-504, 190 U.S.P.Q. 214, 218 (CCPA 1976).
In addition, courts have often combined the issues of lack of utility under 35 U.S.C. § 101, with the asserted absence of an enabling disclosure under 35 U.S.C. § 112, first paragraph. Raytheon Co. v. Roper Corp., 724 F.2d 951, 220 U.S.P.Q. 592 (Fed.Cir.1983). Defendant has similarly asserted that the claimed invention not only does not comply with the enabling requirement of the first paragraph of section 112, but also lacks utility under 35 U.S.C. § 101.10
The Federal Circuit, in discussing the inter-relationship between the enablement requirement and the utility requirement stated:11
To make a claim for a ... process in which these erroneous ideas are incorporated is to stake out a process ... which does not in point of fact exist within the invention. While a patent covering a meritorious invention should not be struck down because the patentee has misconceived the scientific principle of his invention, the error cannot be overlooked when the misconception is embodied in the claim.
Because it is for the invention as claimed that enablement must exist, and because the impossible cannot be enabled, a claim containing a limitation impossible to meet may be held invalid under § 112. Moreover, when a claim requires a means for accomplishing an unattainable result, the claimed invention must be considered inoperative as claimed and the claim must be held invalid under either § 101 or § 112.
In a Court of Claims case also involving the validity of a claimed gyroscope, however, the court stated:12
[T]he law does not require all of the claims to recite each and every element necessary to the operation of the invention. See Brammer v. Schroeder, 106 F. 918 (8th Cir.1901); In re Myers, 410 F.2d 420, 161 U.S.P.Q. 668 [56 CCPA 1129] (1969). Were this not the case, the claims would be prolix to the point of obscuring the inventive concepts to which the claims are to be directed. It is the function of the specification, not the claims, to set forth sufficient detailed information to enable one skilled in the art to make or use the invention, see In re Johnson, 558 F.2d 1008, 194 U.S.P.Q. 187 (CCPA 1977). In the present case, *676these spaces are fully and adequately disclosed in the patent, drawings, and specification ....
Finally, as stated by the CCPA in In re Skrivan, 427 F.2d 801, 166 U.S.P.Q. 85, 57 CCPA 1201 (1970):
[W]e recognize that the disputed limitation deals only with a physical operating condition of an admittedly old process. We see no more reason for requiring that appellant recite the specific angles at which the reactants in his process are to be combined than we do for requiring the recitation of flow rates or size of reactor or any other physical operating condition which might be required in order to obtain an operable process. Those limitations deal with factors which must be presumed to be within the level of ordinary skill in the art. We hold that claims need not recite such factors where one of ordinary skill in the art, to whom the specification and claims are directed, would consider them obvious. [Emphasis supplied.]
Defendant, in asserting invalidity based upon a combination of lack of utility and lack of enablement, first asserts the unworkability of the following aspects of the Barnett gyroscope: (1) translation-rotation pickoff interaction; (2) electromagnetic suspension; (3) radiometer drive; (4) rotor instability; (5) pickoff resolution; (6) rotor-stator gap; and (7) control circuitry. Defendant further asserts that, even assuming these problems can be met, they can only be met through the application of post-filing-date technology. We disagree.
Each and every problem adduced by defendant was straightforwardly solved by plaintiffs, either through their expert witness, Dr. Allen, or through cross-examination of defendant’s expert witness, Mr. Moore. In addition, no substantial evidence was presented by defendant to prove that any of these solutions were necessarily the result of undue experimentation.
Defendant’s only witness, Mr. Moore, presented as a person skilled in the art of gyroscope design, testified to an assortment of alleged shortcomings in the illustrative embodiment presented in the Barnett specification. During his cross-examination, however, Mr. Moore, still testifying as one reasonably skilled in the art of gyroscope design, repeatedly conceded either that he nor anyone else at Rockwell had personal knowledge of any of the asserted problems regarding the Barnett design, or that persons skilled in the art could develop simple solutions to resolve such problems. The evidence adduced at trial makes it clear that problems with the Barnett invention are not beyond the capacity of persons skilled in the art to solve through relatively simple methods and without undue experimentation.
1. Translation-Rotation Pickoff Interaction
Defendant asserts that in the Barnett rotor-position pickoff scheme, with only one axially directed light beam 43, there could be confusion and false readings because photocells 44 would not be able to distinguish between rotor translation, and its rotation vis-a-vis the stator.
On cross-examination, however, Mr. Moore conceded that one skilled in the art could easily solve the problem by having two axially directed light beams emitted on opposite sides of the rotor. Moreover, even with only one light beam directed axially, if the gyroscope’s suspension system were prompt in correcting any translational movement of the rotor from its centered position, any confusion would be of very short duration.
2. Electromagnetic Suspension
Defendant adduces that the use of an electromagnetic field to suspend the rotor, as well as the inhomogeneity of the Barnett rotor will give rise to error torques and variable draft rates.
On cross-examination, however, Mr. Moore stated without hesitation: “I am highly confident that no one at Rockwell has attempted to make the Barnett gyroscope work, even as a study,”; that “No one at Rockwell has attempted to analyze or fabricate the particular embodiment that is shown in the Barnett patent”; and *677that “Rockwell has not proven that the design would not work, that’s correct.” [Emphasis supplied.] Mr. Moore was indulging in conjecture in alleging this problem as proof of the impracticality of the Barnett patent. Moreover, Mr. Moore admitted that the prior art Parker patent, which discloses electromagnetic suspension, describes methods for overcoming these precise difficulties, provided a high degree of magnetic homogeneity is obtained in the rotor. In response to the court’s questions, Mr. Moore declared that one skilled in the art could not overcome such problems with the Barnett rotor design. This should be weighed, however, against his earlier testimony that neither he nor anyone else at Rockwell had even made a study, let alone careful experiments, to see if the Barnett design could work. Mr. Moore also admitted a prejudice against electromagnetic techniques at Rockwell going back to his early days there.
3. Radiometer Drive
Defendant maintains that the radiometric technique for spinning the Barnett rotor requires the presence of some gas, and therefore would not generate sufficient torque, hence sufficient speed for stable operation, because of the gas frictional drag.
When the court asked Mr. Moore for the basis for his allegation, he replied “It’s based — I suppose one could call it conjecture.” [Emphasis supplied.] Mr. Moore also admitted that he had performed no calculations on the subject:13
The Court: But you didn’t prove the unworkability of it through a mathematical formula, is that your testimony?
Mr. Moore: That’s correct.
Moreover, Mr. Moore admitted that he had not conducted even the simplest kind of experiments to confirm his assertions:14
The Court: Have you personally done any experiments plotting out the interference of gas in a gyroscope such as the Barnett gyroscope, say with one percent gas, close to a vacuum, half a percent, five percent, 10 percent, 20 percent, and to see if that was a straight-line proportion or a curve relationship or what relationship?
Mr. Moore: Not exactly as you have related the question, I have not, to determine the quantitative relationship between drag and specific amount of gas present.
The Court: So you don’t know exactly at, say one percent of gas in the vacuum or in the partial vacuum how drastic that detrimental effect would be and whether or not that effect would cause the gyroscope to be substantially inefficient, rather than just very close to efficient.
Mr. Moore: No.
Finally, Dr. Allen, plaintiffs’ expert witness, testified earlier that the radiometric technique would work, even if the gas around the rotor was at atmospheric pressure. Moreover, Dr. Allen’s calculations indicate that with hydrogen around the rotor, the Barnett rotor could rotate with more than sufficient speed. Finally, on cross-examination, Mr. Moore conceded that one skilled in the art might consider utilizing that phenomenon by reducing the gas pressure sufficiently, thus reducing frictional drag, and obtaining high torque by means of large vanes.
4. Asserted Rotor Instability
Defendant avers that the complexity of the construction of the Barnett rotor will cause physical misalignment and differential thermal expansion, as well as spin-axis wobble precluding stable rotation of the rotor. Defendant further asserts there is no provision to damp rotor polhode motion15 or to torque the rotor to the preferred orientation.
On cross-examination, however, Mr. Moore admitted that one skilled in the art could select rotor materials so as to reduce thermal instability. Mr. Moore also agreed *678that the Barnett stator could be turned in an adjustable mounting, so as to be initially aligned with the rotor’s spin axis, no matter how both were then aligned with respect to a chosen inertial reference frame, and that one could then program a computer to compensate for any corrections “for the axis being pointed in a way other than the way you want.” 16 Moore added, that by following the motion of the Barnett axially directed beam, one could compute and compensate for polhode motion as well. This testimony reflected the use of the Barnett gyroscope in the strapdown mode.
Mr. Moore further conceded that speed control is unnecessary, provided the rotor’s speed is known. On cross-examination, he admitted that the Nordsieck patent taught as early as 1961 that the precession speed of a gyroscope rotor is related to its rotational speed; that it can be extracted from the axial rotor-position pickoff signal; that the Barnett rotor would precess; and that, therefore, per Nordsieck, the Barnett rotor speed could be obtained. Mr. Moore further stated that one reasonably skilled in the art would consider such a technique to solve the problem he had identified. Moreover, he stated that each user would have to select a rotor speed in light of the particular application he had in mind.
Finally, it should be noted that Mr. Moore was talking about only the illustrative embodiment, and that he was careful to say only that “one would anticipate a great deal of difficulty in assembling all of these different components in a precisely aligned fashion.” [Emphasis supplied.] On cross-examination, Mr. Moore confessed that in principle, while difficult, it would be possible for one skilled in the art to balance even a complex rotor:17
Mr. Nirmel: Isn’t it correct that you yourself have no personal knowledge and that you are merely conjecturing that there would be a significant unbalance of the rotor that would affect the Barnett gyro operation?
Mr. Moore: That’s correct. I was simply referring, I believe, to the apparent difficulty in achieving an adequate degree of mass balance with such a complex assembly.
Mr. Nirmel: But in principle it is possible to devise techniques to balance complex rotors, is it not?
Mr. Moore: In principle, it is. In practice it has proven more difficult, the more complex the assembly.
5. Rotor-Position Pickoff Resolution
Defendant claims that the large number of components to be carried by the interior surface of the stator, and the concept of an optical photocell rotor-positions pickoff array is impractical and cannot provide sufficient rotor-position pickoff angular resolution to construct a useful gyroscope.
Assuming, arguendo, the truth of the matter asserted, the law is clear that a patented device does not have to compete, on the basis of accuracy, resolution, etc., with later developed more sophisticated allegedly infringing devices. See Decca Limited v. United States, 544 F.2d 1070, 1077, 1080-1081, 210 Ct.Cl. 546, 558, 564-565 (1976). Demonstrating that the 1963 Barnett application may not have met the needs of today’s rotor-position pickoff resolutions, e.g., to the Vaeooth of a degree, or even Vsoth of a degree, does not prove the per se unworkability of the Barnett gyroscope.
6. Rotor-Stator Gap
Defendant argues that the multiple independent optical systems used in the Barnett light beam gyroscope, requires a large rotor-to-stator gap which is contrary to normal gyroscope design practice.
On cross-examination, however, Mr. Moore conceded that the Barnett disclosure does not limit the number of locator beams 26, and that by increasing the number of such beams one could reduce the gap provided one could locate the additional lights 28 and photocells 30 at the stator. He also *679conceded that one skilled in the art might consider making the stator of a transparent material, using fine laser beams, and locating paraphernalia like lights 28 and photocells 30 on the outside surface as a way to effectuate this simple solution to reduce the gap size. Moreover, both the Parker and Nordsieck patents suggest a glass stator that will transmit light and not affect the magnetic suspension field.
Finally, while Mr. Moore was testifying as an expert in gyroscope design in general, it is undisputed that Mr. Moore had never, in his professional career, worked on or with electromagnetically supported rotors in gyroscopes.
7. Control Circuitry
Defendant asserts that there is no control circuitry shown for what is a very complex control situation due to the necessity for interaction among the multiple independent optical systems in the Barnett gyroscope.
While there is no detailed control circuitry disclosed in the Barnett specifications, it should be noted that neither is any detailed control circuitry claimed. The new, useful, and nonobvious teaching of the Barnett invention resides in the active, spherical, free rotor — not in minute details of control circuitry. Defendant did not carry its burden of proof insofar as to demonstrate that such required circuitry was not within the scope of one skilled in the art in 1963.
As a final argument for invalidity, based on a combination of Sections 112 and 101, defendant claims that plaintiffs’ adduced solutions, supra, are not relevant because they are premised on today’s technology, rather than that of 1963. We disagree.
Defendant’s assertions are not persuasive enough to sustain its burden of proof in establishing a prima facie case for invalidity, in view of the presumption of validity the Barnett patent enjoys, particularly because defendant’s sole expert was unsure and hesitant in his responses, even on direct testimony to relevant questions of his own counsel:18
Mr. Daigle: Was the state of fiber optic technology in April 1963 sufficient to achieve that solution, if you know?
Mr. Moore: I do not know, but I do not believe it was.
[Emphasis supplied.]
We do not believe, relying on the wavering type of testimony seen above, that we could find that all the technologies suggested by plaintiffs to resolve the apparent problems asserted by defendant, were post-filing-date in character. We hold, therefore, that with regard to sections 112 and 101, that defendant has not carried its burden of proof as imposed by section 282.
Finally, even if some of defendant’s arguments of unworkability of some aspects of the Barnett gyroscope were true, because they allegedly utilized post-filing-date technology, the defense of non-utility cannot be sustained without proof of total incapacity. E.I. duPont de Nemours & Co. v. Berkley and Co., 620 F.2d 1247, 1260, fn. 17, 205 U.S.P.Q. 1, 10 (8th Cir.1980). Moreover, just some degree of utility is sufficient for patentability. Id. Finally, the fact that an invention has only limited utility and is only operable in certain applications, is not grounds, for finding lack of utility. Raytheon Co. v. Roper Corp., 724 F.2d 951, 958-959, 220 U.S.P.Q. 592, 598 (Fed.Cir.1983); Carpet Seaming Tape Licensing Corp. v. Best Seam, Inc., 694 F.2d 570, 578, 216 U.S.P.Q. 873, 880 (8th Cir.1982).
Unobviousness
While defendant did not specifically address the issue of obviousness, nevertheless, it asserted invalidity, albeit premised on section 112 rather than section 103. Since obviousness is subsumed under validity, and pursuant to the instruction from the Federal Circuit that the lower court “should decide validity and infringement and should enter a judgment on both issues when both are raised in the same proceed*680ing,” 19 out of an abundance of caution, we also address unobviousness.
The applicable statute is, of course, 35 U.S.C. § 103:
A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negatived by the manner in which the invention was made. [Emphasis supplied.]
Some of the more significant words mandated in this statute which delimit the obviousness issue are that we must: (1) determine what “would have been obvious”, as opposed to what “is obvious,” as some courts have been prone to do; (2) address “the subject matter as a whole,” as opposed to selecting individual elements; determining their individual obviousness; and holding the entire invention obvious; and (3) discuss the obviousness of the claimed “invention,” rather than the obviousness of the patent. See Amstar Corp. v. Envi-rontech Corp. et al., 730 F.2d 1476, 221 U.S.P.Q. 649 (Fed.Cir.1984).
The touchstone case which sets forth the guidelines for a correct obviousness determination, is Graham v. John Deere Co., wherein the Supreme Court explicated:
While the ultimate question of patent validity is one of law, the § 103 condition, which is but one of three conditions, each of which must be satisfied, lends itself to several basic factual inquires. Under § 103, the scope and content of the prior art are to be determined; differences between the prior art and the claims at issue are to be ascertained; and the level of ordinary skill in the pertinent are resolved. Against this background, the obviousness or non-obviousness of the subject matter is determined. Such secondary considerations as commercial success, long felt but unsolved needs, failure of others, etc., might be utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented. As indicia of obviousness or nonobviousness, these inquiries may have relevancy. [Graham v. John Deere Co., 383 U.S. 1, at 18, 86 S.Ct. 684, at 694, 15 L.Ed.2d 545 (1966) (Citations omitted).]
Thus, under the Graham analysis, a three-part determination is set forth: (1) the scope and content of the prior art; (2) differences between the prior art and the claim at issue; and (3) the level of skill in the pertinent art. In addition, the Federal Circuit has added a fourth consideration to the above three-part analogy — that of secondary considerations or additional evidence, which may serve as indicia of non-obviousness. In re Sernaker, 702 F.2d 989, 996, 217 U.S.P.Q. 1, 7 (Fed.Cir.1983); Environmental Designs, Ltd. v. Union Oil Co., 713 F.2d 693, 218 U.S.P.Q. 865 (Fed.Cir.1983).20 Other recent Federal Circuit opinions which address the analysis of obviousness and provide a comprehensive guide to proper analysis are: Orthopedic Equip. Co. v. United States, 702 F.2d 1005, 217 U.S.P.Q. 193 (Fed.Cir.1983); Orthopedic Equip. Co. v. All Orthopedic Appliances, Inc., 707 F.2d 1376, 217 U.S.P.Q. 1281 (Fed.Cir.1983); Chore-time Equip., Inc. v. Cumberland Corp., 713 F.2d 774, 218 U.S.P.Q. 673 (Fed.Cir.1983); Carl Schenck, A.G. v. Nortron Corp., 713 F.2d 782, 218 U.S.P.Q. 698 (Fed.Cir.1983).
Moreover, the presumption of validity pursuant to 35 U.S.C. § 282, “obviousness” being subsumed under the issue of validity, applies squarely to the obviousness issue.
*681As Chief Judge Markey stated in Linde-mann:21
Courts are not, of course, at liberty to repeal a statute, or to legislate conditions diminishing its effect. Hence, the statutory presumption cannot “vanish” or be “weakened” and the statutorily assigned burden of proof cannot be shifted. Stratoflex Inv. v. Aeroquip Corp., 713 F.2d 1530, 218 U.S.P.Q. 871 (Fed.Cir.1983). At the same time, much confusion can be avoided by patentees who refrain from efforts to expand the role of the presumption beyond its burden-assigning and decisional approach-governing function.
The burden upon the challenger of validity under 35 U.S.C. § 282 is to introduce evidence of facts establishing invalidity (thus overcoming the presumption). American Hoist & Derrick Co. v. Sowa & Sons, Inc., 725 F.2d 1350 (Fed.Cir.1984). That evidence, if it is to carry the day, must be clear and convincing. Radio Corp. v. Radio Labora-tories, 293 U.S. 1 [, 55 S.Ct. 928, 78 L.Ed. 1453] (1934). Because the mere introduction of non-considered art (a common phenomenon) does not “weaken” or otherwise affect the presumption, there is no basis for adjusting the required level of proof downward to a “mere preponderance.” That the clear and convincing standard may more easily be met when such non-considered art is more pertinent than the cited art means that determination of whether the patent challenger has met its burden turns on the relationship of the uncited art to the claimed invention. Stratoflex, supra.; Railroad Dynamics Inc. v. A. Stucki, [727 F.2d 1506] (Fed.Cir.1984), Solder Removal v. United States Int’l Trade Comm’n., 582 F.2d 628, 199 U.S.P.Q. 129 [65 CCPA 120] (1978).
At trial, however, defendant did not present any evidence to the court to rebut the presumption of validity, as it applies to the issue of obviousness. Defendant did: (1) not present any prior art more relevant than that before the Patent Office; (2) not establish a prima facie case of obviousness; (3) not present testimony by its sole expert on the issue of obviousness, i.e., the four-prong obviousness analysis discussed, supra; (4) not address the obviousness issue in its post-trial brief regarding liability issues; and (5) not propose any findings of fact regarding obviousness. It is therefore clear, a priori, that defendant has not carried its burden, i.e., adducing clear and convincing proof that the claimed Barnett invention, as a whole, as described in claim 7, would have been obvious to one skilled in the art in 1963, the filing date of the Barnett patent application. Because of this void in the evidence of record regarding the obviousness issue, this court need not further address the issues subsumed under obviousness, see Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 1534, n. 3, 218 U.S.P.Q. 871, 875-76, n. 3 (Fed.Cir.1983), but only need find, as we do, that the defendant did not carry its burden of establishing invalidity, through obviousness, of claim 7 in suit, as required by 35 U.S.C. § 282. Id.
Conclusion
Claim 7 of Barnett U.S. Patent No. 3,379,889, entitled “Beam-Driven Gyroscope Device” is valid, but not infringed, either literally or under the doctrine of equivalents.
At the conclusion of plaintiffs’ case, defendant moved for dismissal pursuant to RUSCC 41(b) on the ground that plaintiffs failed to establish a prima facie case of infringement. The court decided to hear all the evidence in the case before rendering a judgment. On the basis of all the evidence before the court and in accordance with this opinion, plaintiffs’ complaint is to be dismissed.
FINDINGS OF FACT
A. Background Of The Case
1. This is a patent infringement suit brought under 28 U.S.C. § 1498. Plaintiffs *682seek reasonable and entire compensation for the unauthorized manufacture, or use by or for the United States, of the invention described and claimed in U.S. Patent No. 3,379,889, titled “Beam-Driven Gyroscope Device”. The patent was issued on April 23, 1968, on an application filed April 5, 1963, by coinventor-plaintiffs Eugene R. Barnett and his since deceased father, Willard L. Barnett. Plaintiff, Eugene R. Barnett, resides at 6268 Windsor Drive, Indianapolis, Indiana. (PX-1)
2. Plaintiffs are litigating only Claim 7 of the Barnett patent in this suit. Only the issues of validity and infringement of claim 7 of the patent are before the court at this time. A trial on the merits was held specifically addressing the issues of validity and infringement of claim 7 of the patent in suit. The accounting issue has been deferred for a later trial pending determination of liability. (See Memorandum of Preliminary Conference, filed August 10, 1981.)
History of the Gyroscope
3. The basic principle, upon which gyroscopes are based, is that a rotating body tends to maintain a stable orientation in space. (DX-39, pp. 525-526) In the mid-nineteenth century, the French scientist J.B.L. Fouchault constructed a gyroscope in which the rotor was mounted in three supporting frames known as gimbals in which it moved freely, and demonstrated that a rotor maintained its original orientation with respect to inertial space, and did not follow the earth’s rotation. Fouchault was also responsible for naming the device a gyroscope. (DX-39, p. 525; Moore Tr. 405) While Fouchault’s experiments suggested that the gyroscope could be used to indicate direction, the first practicable gyroscope compass was not developed until the early 1900’s when it was used for marine applications. (DX-39, p. 525; Moore Tr. 405-406)
4. Until World War II, gyroscopes were usually gimballed devices in which the rotor was mechanically attached to its gimbals through bearings. However, the friction attributal to the bearings caused the rotor to precess or drift from the stable position. (DX-36, p. 6) Precession is attributable to the physics of angular momentum whereby a torque acting at right angles to the spin axis of a rotating body causes the axis to move at a right angle to the force. (DX-39, p. 526; Moore Tr. 444) Gyroscopes of the 1940’s used jeweled bearings and typically had drift rates of the order of one to ten degrees per hour. (Moore Tr. 406, 407; DX-36 pp. 6, 9, 10) The development of the ballistic missile by Germany near the end of World War II, and its improvement through the 1950’s into a missile having intercontinental range as well as the adaptation of the nuclear submarine to serve as a mobile sea-based launching platform for ballistic missiles, created a need for gyroscopes with much lower drift rates of the order of one-tenth or one-hundredth degree per hour to be used in the inertial navigation systems of these missiles and submarines. (Moore Tr. 407-410; DX-36, pp. 5-13; DX-37, p. 48)
5. A basic principle of an inertial navigation system is the computation of a vehicle’s present position by monitoring acceleration forces acting on a navigation instrument carried by it. (DX-36, p. 2; DX-39, pp. 527-528; Moore Tr. 409-410) The navigation device is first stored with initial position information, and the changes in the position of the vehicle are constantly computed by performing a double integration with respect to time of the acceleration forces sensed by the instrument. (DX-36, pp. 2-3) Acceleration forces are usually sensed by three accelerometers mounted in orthogonal relationship upon a stable inertial platform. The platform is maintained in a stable relationship to space or to the surface of the earth by one or more gyroscopes. (DX-39, p. 527; Moore Tr. 420-422)
6. In an effort to minimize the drift rate in gyroscopes, efforts were made to reduce gimbal friction or to eliminate the gimbal friction or to eliminate the gimbal bearings altogether. During the late 1940's and the early 1950’s, floated gyroscopes were con*683structed whereby the gimbal assembly was “floated” in a liquid or supported by a current of air to reduce the load carried by the gimbals. (DX-36, pp. 6, 9, 10, 94-96, 100; DX-37, pp. 48, 70, 71) The air or liquid floated-bearing principle was also applied to the rotor itself to eliminate the need for a mechanical bearing between the rotor and stator of the gyroscope. (DX-37, p. 72)
7. In the early 1950’s, electrically-suspended gyroscopes were developed by Arnold Nordsieck at the University of Illinois and Jesse Beams at the University of Virginia. (Allen Tr. 120-122; Moore Tr. 410-412) Dr. Nordsieck’s device used an electrostatic field to suspend the gyroscope rotor and is described in his U.S. Patent No. 3,003,356, which was filed on November 1,1954. (DX-20; Allen Tr. 120; Moore Tr. 412) Jesse Beams’ gyroscope used an electromagnetic field for this purpose and is described in U.S. Patent No. 2,691,306, filed on January 30, 1951. (DX-18; Allen Tr. 120) Hence, well before 1963, it was known in the art that a free rotor gyroscope could be constructed without gimbals or other mechanical bearings by suspending a gyroscope rotor with a fluid or gas bearing, or by an electrostatic or electromagnetic field.
8. As previously noted, the property which makes the gyroscope a useful device is the tendency of the angular momentum of the spinning rotor to maintain the orientation of the spin axis of the rotor fixed in relation to inertial space. Consequently, it is necessary to be able to ascertain the angular orientation of the rotor spin axis, a function which is known in the gyroscope art as the rotor-position pickoff scheme. (Allen Tr. 229-231) In a free rotor gyroscope, in which mechanical bearings have been eliminated through the use of a fluid or gas bearing or an electric field to suspend the rotor, it is undesirable to require any mechanical connection to the rotor (from the stator) to determine the spin axis orientation. This led to the concurrent development of optical and electrical systems for detecting changing rotor-spin axis orientation in free rotor gyroscopes. Nord-sieck utilized the capacitance effect of a capacitive bridge, formed by electrodes on the stator and a protuberance on the rotor, to sense rotor orientation. (DX-20, col. 4, lines 3-36; line 59 through col. 7, line 34; Fig. 6; Moore Tr. 614; see also DX-19, DX-37, p. 72; Moore Tr. 614-618) Rotor-position pickoff schemes, utilizing the electromagnetic inductance effect, were also known in the art prior to 1963. (DX-23, 32, 34, 35) Similar optical schemes, utilizing photoelectric pickoff sensing, were well known in the art by that time. (DX-8, 14, 33)
9. Vehicles utilizing gyroscopes include aircraft, ballistic missiles, surface ships, and submarines. The accused gyroscopes are manufactured for defendant by the Rockwell International Corporation. (Moore Tr. 415-417; Moore Tr. 422-425)
B. Gyroscopes Generally, and Definitions
10. A “gyroscope ” is a device in which one of its elements, the “rotor", rotates about an imaginary axis, generally referred to as the rotor’s “spin axis ” going through the rotor’s center of mass. That portion of the gyroscope which is immediately adjacent and exterior to the spinning rotor is generally referred to as the gyroscope’s “stator". (DX-39)
11. The rotor’s positional “stability” depends on its angular momentum, the product of its rotational moment of inertia and the angular velocity. Therefore, the greater the rotor’s inertia or its rotational speed, the more stable is its spin axis. (Allen Tr. 164; Moore Tr. 722, 262) When an external torque acts on the spinning rotor at an angle to its spin axis, the spin axis is deflected and tends to rotate about an axis normal to the external torque vector; this motion is called “precession ” of the rotor’s spin axis. (Moore Tr. 444, DX-39) The rotor’s performance, however, is independent of the direction in which it rotates. (Allen Tr. 163)
12. It has long been the goal of gyroscope designers to support the rotor inside a stator in a way that obviates external *684torques acting on the rotor, and hence eliminating precession of the rotor’s spin axis. One obvious problem in early gyroscope designs was the frictional torque at the rotor bearings. As indicated supra, the solutions tried included liquid or gas floatation of the rotor, as well as electromagnetic or electrostatic force fields strong enough to levitate or suspend the rotor, with no physical contact between it and the stator. (DX-36, 37, 38 and 72) Both these suspension modes, i.e., electromagnetic and electrostatic, require rotors made of electrieal-conducting materials, in which eddy currents are generated as the rotor spins. These eddy currents then interact with the causative external magnetic fields, e.g., the earth’s own magnetic field, tending to slow the rotor’s spin thereby causing precession. (Allen Tr. 220-224) Various techniques, such as Rockwell’s mumetal shield, have been tried to eliminate the influence of such extraneous magnetic fields and the resultant precession-causing torques. (DX-46B and DX-31, Figure 2-2-8, display the Mumetal Shield used by Rockwell.) (See infra.)
13. Even so, gyroscopes intended for prolonged use are designed so that energy can be supplied via an externally applied torque to counter the drag torques experienced by the rotor from residual air or gas friction, or due to the influence of extraneous magnetic fields. Depending on how this is effected, and on the rotor’s size and *685mass, this replenishment or energy may be_ done intermittently or continually. (PX-1, Figs. 1, 5, infra; PX-1, Col. 2, In. 52 — Col. 3, In. 2; PX-2; Allen Tr. 137-138; Allen Tr. 198-199; DX-31)
14. The accuracy of a gyroscope is char-aeterized by its “drift rate ”, measured in degrees per hour, the rate at which its rotor’s spin axis drifts away from a partic*686ular orientation in the inertial reference frame under the action extraneous torques. The smaller the drift rate is, the more stable the gyroscope. Drift rates of about 1 degree per hour were achieved by the end of World War II and current capabilities are better than 0.01 degree per hour. (Moore Tr. 406-407) It is accepted practice to ensure a stable and measurable drift rate if it cannot be ignored for accurate missions, and then to compensate for it in use. (Moore Tr. 450) This may necessitate knowing or controlling the rotor’s angular speed. (Moore Tr. 486) The amount of the angular error, once knowing the rotor’s initial spin axis position, depends on the time integral of the drift rate. (DX-37) An inertial navigation system can be considered to be a special case of a dead-reckoning navigation system in which the errors grow with time because of the drift-rate of the gyroscope. The user’s desire is to keep such errors as low as possible, particularly for long-range missions, e.g., in nuclear submarines on patrol, in order to avoid requiring the vessel to expose itself when obtaining an external position fix. (Moore Tr. 409-410)
15. “Inertial Navigation ” of the vehicle carrying the gyroscope, requires that the vehicle’s motion be purposeful in as many as all three spatial dimensions and in time. In practice, this generally requires knowledge of where the vehicle started its motion, how fast and in which direction it is moving at any instant, the magnitude and direction it is moving at any instant, and the magnitude and direction of its instantaneous acceleration at any moment. The gyroscope, with separate accelerometers (Rockwell), or by itself (Barnett), provides the instantaneous magnitude and direction of the vehicle’s acceleration vector as it moves. (Moore Tr. 420) Knowing both the direction and magnitude of his translational acceleration at each of two successive instants in time, the user will integrated the acceleration vector over time; once, to determine his instantaneous velocity vector, and then a second time to determine his instantaneous position, i.e., exactly where he is with respect to his starting point. Once knowing the desired direction, and the particular route, one can chart a course accordingly, monitor the progress and made course corrections. In practice, this is done on a computer because of the large volume of data and calculations involves. (DX-39, pp. 527-528)
16. The gyroscope’s spinning rotor provides a reference direction either constant in time or one whose drift is known and compensated for. (Moore Tr. 450) In addition, the magnitude of the vehicle’s translational acceleration must be measured in three mutually orthogonal directions whose orientation with respect to the rotor’s spin axis is known. Some gyroscopes have so-called “stable platforms”, which orientation is kept fixed with respect to the stable rotor axis (or axes when more than one gyroscope is utilized simultaneously), to carry three mutually orthogonal accelerometers. (Moore Tr. 420) The preferred embodiment discussed in the Barnett patent teaches an alternative way of measuring acceleration both in magnitude and direction, in the normal course of operating the gyroscope, without separate accelerometers outside the gyroscope itself. (PX-1, Col. 2) However, Claim 7, the only Barnett claim in issue, is not concerned with the specific method by which translational acceleration is measured.
17. The stator cavity usually is similar in shape to the rotor contained within. The Barnett rotor, when it spins, has an essentially spherical shape. (PX-1, Fig. 1) The Rockwell rotor is spherical in shape but describes a slightly torodial path because its center of rotation does not coincide with its geometric center. Both the Barnett and Rockwell gyroscopes have spherical stator cavities. (PX-1, Fig. 1; DX-31, Fig. 2) In normal operation, the rotor’s center of rotation (usually also its center of mass) is maintained at the stator’s geometric center regardless of the rotor’s shape. Any translational deviation from this position must be rapidly determined and promptly corrected. This “centering” of the rotor is done by means located on the stator capable of sensing the rotor’s spin-axis position, *687and actuating the translational corrective response, to return the rotor to its specified position, by applying a carefully monitored force from the stator. This “suspension force ” overcomes gravitational and translational accelerations experienced by the rotor. “Newton’s Second Law of Motion ” requires that the force applied by the stator to the free rotor be directly proportional to the relative acceleration between them. This statement puts into words the basis equation of Newtonian mechanics:
F = M x a
(force) = (mass) x (acceleration)
18. There are two ways of supporting the gyroscope’s stator in the user vehicle: first, by having it fixed rigidly to the vehicle (a strapdown gyroscope) or, second, by having it carried on one or more gimbals (a gimballed gyroscope). (Tr. 417)
(a) The “strapdown gyroscope” needs to have no moving parts other than its rotor and may, therefore, be relatively sturdy, simple and inexpensive. As the used vehicle moves about, the stator of the strapdown gyroscope moves with it vis-a-vis the rotor’s stable spin axis. Thus, the motion of the stator, measured with respect to a reference direction related to the rotor’s axial position represents the motion of the vehicle itself. (Moore Tr. 417)
(b) A “gimbal ” is probably best visualized in its simplest form to be like a universal joint with two bearing axes orthogonal to each other such that a body, e.g., a gyroscope stator, can turn about the axes through those bearings. A complete set of such assemblies would provide for complete isolation of anything carried inside the gimbals from the external motion of the user vehicle. (Moore Tr. 418; DX-31)
19. For prolonged missions, e.g., on nuclear submarines that stay submerged for months at a time, a “gimballed gyroscope” supported so as to have four degrees of freedom (one being redundant), with servomotors controlling each gimbal, is preferred over the strapdown type. (DX-31, 45; Moore Tr. 415, 418) Deviations from the specified relationship between the stator and the rotor’s axial position, generally referred to as an “error function” or “error signal” are utilized to manipulate the gimbal servos for corrective action. (Moore Tr. 421-422) A gim-balled gyroscope is capable of much greater accuracy than the strapdown type. (Moore Tr. 419) The user actually measures the relative change in angular position between the stator and the vehicle, since the stator’s orientation is essentially locked on to that of the stable rotor spin axis.
20. A gyroscope capable of functioning in the strapdown mode can be utilized in the gimballed mode when it is mounted on proper gimbals. (Moore Tr. 415-416) The reverse is not true; a gyroscope designed so that the rotor and stator must maintain very closely a specified orientation with respect to each other cannot function as a strapdown gyroscope. The specific gyroscope embodiment disclosed in the Nord-sieck patent cannot be used as a strapdown gyroscope. (DX-20, Fig. 2; Moore Tr. 741-742) The Rockwell gyroscopes, when used in the gimballed mode, are firmly affixed to a stable platform which carries three mutually orthogonal accelerometers to measure the three components of translational acceleration in the inertial frame. (Moore Tr. 420-422) The stable platform carrying the accelerometers, therefore, has its angular position unchanging in the inertial reference frame.
Prosecution History Estoppel of the Barnett Patent
21. The application for the patent in suit was filed on April 5,1963, by plaintiffs Eugene R. Barnett and his father Willard L. Barnett, the latter since deceased. (DX-1) The application described a gyroscope device in which the rotor was suspended by an electromagnetic field and light beams were used to fulfill several other necessary gyroscope functions. (DX-1, pp. 1-8) The device is best shown in Fig. 1 of the patent, supra. The gyroscope consists of a generally circular rotor 10 mounted within a stator 12 attached to a housing 14. (PX-1, *688col. 1, line 70 through col. 2, line 5) Electromagnets 22 set up an electromagnetic field which suspends the rotor in the center of the stator. (PX-1, col. 2, lines 23-28; Fig. 4; Allen Tr. 137; Moore Tr. 434; DX-41 red items) The translational position of the rotor within the stator is monitored by a photoelectric system consisting of lights 28 projecting light beams 26 onto photocells 30. (PX-1, col. 2, lines 29-36) Movement of the rotor from the desired position alters the pattern of light falling on the photocells which send signals through control device 18 to regulate the electromagnets to return the rotor to the desired position. (PX-1, col. 2, lines 36-44; Fig. 3; Allen Tr. 135-137; Moore Tr. 472-473; DX-41 orange items) Another set of light beams 36 is created by lamp array 34 to cause the rotor to spin. (PX-1, col. 2, lines 53-58) This is accomplished by a set of recessed radial vanes 38 mounted along the equator of the rotor which intercept light beams 36 and provide a rotational force on the rotor through the radiometer effect. (PX-1, col. 2, lines 59-66; Fig. 5; Allen Tr. 138; Moore Tr. 477-478; DX-41 green items) The radiometer effect is caused by the unequal rebound of the residual gas molecules in the stator off the dark side of the vanes which are heated by the energy of the light beams. (PX-1; col. 2, line 68 through col. 3, line 2; Allen Tr. 138-168, 252-260; PX-2; DX-40) In order to determine or to “pickoff” the information concerning the angular orientation of the rotor’s spin axis, a lamp 40 and focusing lens 42 are mounted at one end of the rotor spin axis to project light ray 43 onto photocell array 44 mounted on the stator. (PX-1, col. 3, lines 6-14; Fig. 7; Allen Tr. 168-169; Moore Tr. 545; DX-41 blue items) Lamp 40 is powered by an active energy conversion system mounted on the rotor in which the radiometer vanes 38 also contain photovoltaic cells 47 which receive energy from the rotor drive light beams 36 and convert this light energy into an electrical current which flows through wires 48 to power the lamp. (PX-1, col. 3, lines 14-18; Fig. 6; Allen Tr. 168-169; Moore Tr. 545) The orientation of the rotor’s spin axis is determined by the position at which light ray 43 falls upon the photocell array 44. (PX-1, col. 3, lines 19-30; Allen Tr. 167-171; Moore Tr. 545)
22. In the first Office Action dated April 7, 1964, the Patent Office Examiner cited four United States patents as prior art and rejected all eleven claims in the Barnett application. (DX-1, pp. 24-25) Wittkuhns U.S. Patent No. 1,999,646, An-nen U.S. Patent No. 2,378,744, and Dias U.S. Patent No. 2,541,217 were each cited as disclosing the use of magnetic means to locate a gyroscope member in a stator with an energy source and conversion means to impart rotation to the gyroscope and sense its direction of movement. (DX-1, pp. 24-25) More specifically, Wittkuhns and An-nen disclose gyroscopes in which the movement of the rotor’s spin axis is detected by a photoelectric system which maintains the desired orientation of the rotor’s spin axis is detected by a photoelectric system which maintains the desired orientation of the rotor within the stator. (DX-3, 4) Dias discloses the use of a photoelectric system to determine the position of a compass needle in an automatic ship steering system. (DX-5) Hammond teaches the use of a photoelectric system mounted on the rotating member of a steering control system in which the rotating member tracks the position of a light source. (DX-2)
23. In an amendment filed October 7, 1964, applicants argued that the references cited by the Patent Office did not disclose the use of a magnetic means to locate the gyroscope member, as the Examiner had asserted, and that the reference also failed to disclose several specific features of the applicants’ gyroscope. (DX-1, pp. 26-37) Of particular relevance to this case, applicants stated in reference to application claim 8 (which eventually issued as patent claim 7 in suit) that the references failed to show the use of an energy conversion means located on the rotor which converts some of the light used to spin the rotor into electrical energy to power the light which projects the spin axis light ray 43 from the rotor to the stator. (DX-1, pp. 34-35) In *689response to these arguments, the Examiner issued a second Office Action dated December 28, 1966, in which ten prior art patents were cited to reject all claims as obvious under 35 U.S.C. § 103. (DX-1, pp. 38-40) The principal reference was Parker, U.S. Patent No. 2,919,583 which also discloses a magnetically supported gyroscope. (DX-8) Figure 1 of the Parker patent is reproduced, infra.
24. The Parker gyroscopes consists of a spherical rotor 10 mounted in an enclosure 25 and suspended by an electromagnetic field created by electromagnets 11, 12, 13, and 14. (DX-8, col. 3, lines 37-43) The translational position of the rotor is maintained by a photoelectric control system comprised of lamp 18 and photoelectric cell 22 in essentially the same manner as the Barnett gyroscope. (DX-8, col. 3, lines 43-60; col. 3, line 68 through col. 5, line 3) The rotor is driven by an electron gun 27 which projects a beam of electrons 28 tangentially on the surface of the rotor. (DX-8, col. 5, lines 9-26) The position of the rotor’s spin axis is determined by photoelectric detector 71 which detects the modulation pattern of the beam of light projected by lamp 18 upon the rotor and reflected by a band having different optical properties placed on the surface of the rotor. (DX-8, col. 5, lines 27-60)
*690
25. In this second Office Action, Beeh U.S. Patent No. 3,268,735, was cited as showing the Barnett radiometer drive system as an obvious substitute for Parker’s electron gun. (DX-1, p. 40; DX-15), Simon, U.S. Patent No. 3,225,608, and Elwell et al., U.S. Patent No. 3,254,537, were cited as showing the use of a photoelectric system to sense and control the angular orientation of the spin axis of a rotor. (DX-1, p. 40; DX-12, 14) Marrison, U.S. Patent No. 2,919,358, was cited as showing the use of photovoltaic cells to generate current within a rotor without mechanical connections. (DX-1, p. 40; DX-7)
*69126. Applicants responded to the second Office Action in an amendment filed March 27, 1967. (DX-1, pp. 41-50) In this amendment, the applicants explicitly distinguished their invention from the gyroscope shown in Parker by focusing the Examiner’s attention to the double function performed by Barnett’s energy beams 36 which provides both the imparting rotation of the rotor, and the energizing of the rotor-circuit 48, which energizes energy-means 40, which projects energy ray 43 for the rotation-sensing cells 44. In contrast, the primary reference Parker, uses an electron beam 28 to impart rotation to its rotor 10, but does not use the electron beam for the second purpose as stated by the Barnett coinventors. (DX-1, pp. 43-44) On June 23,1967, the Examiner allowed claims 1-3, 5-10 and 12 “in view of applicants’ persuasive and clearly presented arguments.” (DX-1, pp. 51-52) Application claim 8 became claim 7 in the patent in suit which issued on April 23, 1968. (DX-1, pp. 15-17; PX-1)
27. Claim 7 in suit, presented in subpar-agraph form, is as follows: (PX-1)
Claim 7
(a) A control means, comprising:
(b) a stator means
(c) and a movable means;
(d) locating means for locating said movable means in a certain location with respect to said stator means;
(e) first energy means, responsive to movement of said movable means away from said certain location, for energizing said locating means;
(f) second energy means for imparting rotation to said movable means; the said movable means being rotatable in response to energy from a second energy means;
(g) third energy means carried by the movable means; the said third energy means being actuated by said second energy means;
(h) indicating means carried by the stator means, responsive to energy of said third energy means, to indicative relative movement of the axis of the movable means with respect to the stator means;
(Claim element (f) is rearranged from the order in which its two parts appear in the patent.)
28. Plaintiffs did not present any evidence at trial to establish a date of invention prior to April 5, 1963, the filing date of the patent application, and stipulated that this is the date of invention for purposes of this litigation. (Maxwell Tr. 44) Plaintiffs did not call Eugene R. Barnett, the surviving inventor and one of the plaintiffs, as a witness at trial, nor was he in attendance in Los Angeles during the trial. (Maxwell Tr. 6, 7, 32-46) Plaintiffs stipulated to the admission of a portion of Mr. Barnett’s deposition testimony in which he stated that patentees have never constructed a complete working embodiment of the gyroscope described in the patent in suit. (Tr. 807-808)
Relevant Field of Art: Gyroscopes Containing Spherical Free Rotors
29. While it is recognized that “active ” spherical free rotors do not constitute a recognized field of art, nevertheless, plaintiff asserts that there are only two gyroscopes that have active spherical free rotors: the patented Barnett invention and, as described more fully below, the accused Rockwell gyroscope. (Allen Tr. 347-348; Moore Tr. 745-748) The accused Rockwell ESG’s rotor, however, rather than having an active rotor as asserted by plaintiff, is more properly described as an inactive rotor, or passive rotor. It is in fact merely a passive element which possesses energy which is impressed upon it. (Moore Tr. 792-794, particularly lines 6-12 at Tr. 794) Knowledge of the orientation of the rotor’s axis vis-a-vis the stator is absolutely essential for navigation. Various techniques have been proposed, tried, and used over the years to obtain this information, often referred to as the determination of “angular pickoff” or to determine a “rotor’s *692spin axis position”, accurately. (DX-8, 31, 33; PX-1; Allen Tr. 230, 326)22
30. Where the rotor plays merely a “passive” role in the process, e.g., by reflecting light from its distinctively marked surface to observation points on the stator, the rotor does not expend any of its own energy to provide axial position information. Excellent examples of gyroscopes using passive rotors, are those taught by Kunz and Parker. (DX-8, 33; Moore Tr. 747-748)
31. On the other hand, where the gyroscope has an “active ” rotor, i.e., one that must receive, convert, and send that third energy to transmit its axial position information to the stator, it becomes necessary to continuously replenish this energy of the rotor from the stator for prolonged accurate performance of the gyroscope. Since no evidence was adduced at trial that the patented Barnett gyroscope was not the first to employ an “active ” spherical free rotor in a gyroscope, we assume for the purposes of this trial, that the patented Barnett gyroscope was the first, as asserted by plaintiff. The accused Rockwell gyroscopes, on the other hand, do not utilize the “heat”, which is generated by the rapid redistribution of its electrical charges on its rotor surface, to transmit its axial position information to the stator. (Allen Tr. 326-327. See also infra, FF’s 61 and 76)
Teaching of the Barnett Patent
32. The Barnett invention relates “particularly to control means utilizing gyroscope components for control of vehicles.” (PX-1, Col. 1) The coinventor-plaintiffs patented a “new and improved gyroscope control means having advantageous features of sensing, operation and responsiveness.” (PX-1, Col. 1) Without using language limiting the practice of their invention to any particular means, the plaintiffs explain in their patent specification, as follows, how cooperating elements in their illustrative embodiment would perform the three essential functions of a useful gyroscope. (PX-1, Col. 2)
The various concepts and features of the Barnett invention are presented in the patent specification under these headings:
(a) Means for locating the rotor in the stator, and for sensing and correcting location-deviation of the rotor;
(b) Means for imparting rotation to the rotor; and
(c) Means for sensing and correcting relative rotation of the rotor and stator.
(PX-1, Col. 2)
33. The Barnett patent contains a “ball-like” essentially spherical free rotor which in use is suspended under the action of forces exerted on it from the essentially spherical stator surrounding it. (PX-1, Col. 1) The term “essentially spherical” means that while the shapes referred to possess overall sphericity of geometry, both the Barnett rotor and stator may have recesses within them to accommodate various elements to perform specific functions. (PX-1, Fig. 1-17)
34. In the illustrative embodiment, the Barnett patent discloses electromagnets in the stator, capable of generating a composite electromagnetic force field within the stator cavity, to exert a net force on magnetic material in the rotor to suspend or levitate it as a free rotor, under all operating conditions. (Allen Tr. 122-125, 127-128) These are electromagnets identified as 22 in Figures 1, 2 and 4 of the Barnett patent. (PX-1) They are not permanent magnets each of a fixed strength but, by design, individual electromagnets along each geometric axis, exerting precise control from all directions to counter gravitational and translational accelerations to maintain the rotor at the center of the stator cavity. The Barnett patent shows a control unit 18, to exercise, inter alia, this suspension/location control over the elec*693tromagnets 22. (PX-1, Figs 1, 4, Col. 2; Allen Tr. 137) There is nothing specific about the Barnett embodiment that requires only an electromagnetic suspension; an electrostatic drive could be utilized in the Barnett gyroscope, assuming the separation between the rotor and stator was made substantially smaller; the achievement within one skilled in the art; and if there were no preclusion due to prosecution history estoppel.
35. The Barnett patent teaches that the energy required to center the rotor, may be used to determine the user vehicle’s translational acceleration. (PX-1, Col. 2; Allen Tr. 669-670)
36. In its illustrative embodiment, the Barnett patent suggests a multiplicity of narrow light beans 26, projected from sources 28, on the stator so as to very closely bracket the spining rotor H) and reach light sensitive photo-electric cells 30 connected to the control unit IB. (PX-1, Figs. 1, 3) When the rotor moves, relative to its designed-eentered position inside the stator, it blocks one or more of these light beams 26. This affects corresponding photoelectric cells 30 and provides error signals to control unit 18, thereby causing it to adjust the suspension forces exerted by suspension electromagnets 22. (Allen Tr. 135-137, 472) Thus, the Barnett patent shows in its illustrative embodiment a photoelectric rotor location system that works cooperatively with the electromagnetic suspension system to correct rotor-to-stator spin-axis deviation, while the latter system simultaneously measures the vehicle’s translational acceleration. (PX-12; Moore Tr. 767-772)
37. The Barnett patent, in its illustrative embodiment, does not suggest that the figures are drawn to scale, nor does it specify any particular magnitude for the stator or rotor geometries, rotor speed, number or size of rotor vanes, electrical power to the suspension electromagnets, or sensitivity for the photoelectric cells. (PX-1, Col. 1; Allen Tr. 150-151) Persons skilled in the art, seeking to apply its teaching, presumably would select such parameters to suit the projected use or mission at hand, and may readily adjust the number and direction of rotor-locating light beans to minimize the radial gap between the rotor and stator. (PX-12; Moore Tr. 764-771) Likewise, for different purposes, e.g., rotor spin-axis location or rotor location sensing, lights of different frequencies may be used to avoid interference among different light-sensitive elements as they function cooperatively. (PX-1, Col. 3; Allen Tr. 178)
38. Once the Barnett free rotor is levitated and centered, it must be spun up to a useful speed, i.e., it must be given sufficient angular momentum to provide a stable axial position for reference by the user. (Allen Tr. 263-264; PX-5, 8; Moore Tr. 722-728) However, despite substantial evacuation of the stator cavity, there will remain some low pressure gas around the rotor, exerting a frictional drag which would in the time slow it down to an unacceptable low speed. Since its rotor is likely to contain at least some electrically conducting material, there may also be eddy currents generated therein by extraneous magnetic fields, e.g., the earth’s own magnetic field, which may lead to interactions causing additional drag torques slowing down the rotor. (Allen Tr. 224-226) Hence the rotor’s kinetic energy loss must be continually replaced by applying an external torque to it.
39. Recognizing that the rotor must be spun up, and its kinetic energy thereafter replenished to overcome drag torques, the Barnett patent, in its illustrative embodiment, suggests a solution that requires neither magnetic forces nor a mass transfer from the stator to the rotor, to achieve and maintain rotor spin. (PX-1, Figs. 1, 5; Allen Tr. 138; PX-1, Figs. 1, 5) The suggested technique involves certain energy sources, i.e., lights 34 in Barnett Patent PX-1, Figures 1 and 5, projecting energy 36 from the stator 12 to the spinning rotor 10. The amount of energy so projected would be regulated by control unit 18. (PX-1, Col. 2) The rotor would have around its “equator”, i.e., in the plane nor*694mal to the spin axis, a set of vanes, each colored dark on one side and light on the other. Persons skilled in the art could shape the vanes so as to reduce the gas frictional drag. (Allen Tr. 266-269; PX-5)
40. In one scenario, the dark face of each vane would absorb more of the incident energy, and be at a higher temperature than its opposite lighter-colored and more reflective side. When the residual gas molecules, still present after substantial evacuation of the stator cavity, bounce off the darker side of each vane, they do so at a higher velocity than they do from the lighter side. (Allen Tr. 142-144) As a result, per Newton’s Third Law of Motion, as they leave the dark side of each vane, they impart to it a greater reactive momentum than they do on the light side. (DX-40; Allen Tr. 257-258) This disparity in the net force on each of many vanes, acting at a distance from the rotor’s spin axis, generates a torque capable of speeding up á rotor from rest, or of replenishing the rotor’s kinetic energy of rotation in spite of drag torques. This effect, generally referred to as the “radiometric effect’’, requires the presence of some gas around the rotor. The Barnett rotor would turn by this means, even with air at atmospheric pressure around it, albeit slowly. (Allen Tr. 145) However, one skilled in the are may consider using low pressures and various types of air molecules of low molecular weight, e.g., hydrogen, to increase rotor spin speed significantly. (PX-5) The darker vanes in this scenario will always move away from the incident light, thus deciding the direction of the rotor’s spin. (Allen Tr. 144-145; 154-156)
41. In an alternative scenario, even if all the gas molecules were removed from the stator cavity, one skilled in the art could still cause the Barnett free rotor to rotate sufficiently to function as a gyroscope. The photon theory of light, which is an alternative to the wave-theory of light, provides the explanation for such a rotation, as follows: photons are discrete quanta of light energy, which reach both light-colored and dark vane surfaces with equal incident momentum, and are reflected back with essentially reversed momentum at the lighter vane surfaces, but are simply absorbed at the darker surfaces. The consequence is that, for each incident photon, the lighter surface, reflecting it back, reverses the photon’s momentum, and therefore itself experiences an impulsive reaction force, per the law of physics, equal to twice that experienced, for each photon received and absorbed, at that vane’s dark surface. (PX-1; Allen Tr. 159-161, 254-255; Moore Tr. 681) The fact that the dark surface absorbs more energy, makes it warmer than the light surface; but since it is assumed that no gas is present in this scenario, there is no gas friction drag torque to impede the rotational motion of the rotor. (Allen Tr. 165-167) Overall, given uniform lighting around the rotor in this scenario, the rotor will turn so that the lighter surfaces move away from light incident on them. This is the exact opposite of what happens to the direction of rotation, when the rebound of gas molecules spins the rotor. (Allen Tr. 165-167)
42. Person skilled in the art of gyroscope design, could reasonably be expected to consider both alternatives. (Allen Tr. 252-255) In practice, it would be impossible to extract every single gas molecule from the stator cavity. Nevertheless, when a strong enough vacuum is obtained, i.e., the number of gas molecules is sufficiently reduced, the operation of the rotor would shift from scenario one to scenario two; and the Barnett gyroscope would work equally well with the rotor spinning in either direction. (Allen Tr. 160-161, 165; PX-2) The incident light energy does not have to have a preferred orientation either; uniform illumination around the stator will work perfectly well and, in fact, would be necessary for strap-down gyroscope operation. (Allen Tr. 163; PX-11; Moore Tr. 755, 763)
43. Energy inputs to the stator, no matter how caused, would heat up the gyroscope unless proper cooling is provided. The Barnett specification allows for temperature control in element 20, regulated by control unit IB. (PX-1, Fig. 1, Col. 3) *695Since there is no disclosure regarding element 20, the patentee apparently assumes it to be within the technology of one skilled in the gyroscope art.
44. In its illustrative embodiment, the Barnett patent discloses photovoltaic cells 46 integral with the rotor and distributed among the radiometric torquing vanes 38 at the equator of the rotor, wire 38 to carry the electrical current generated by these cells to a light bulb 40, the light therefrom being focussed by lens 42, as a fine light beam 43 along the rotor’s spin axis, to receptive photo-sensitive cells 44, distributed about the stator’s inner surface. As the stator turns vis-a-vis the stable spinning rotor, the axially-directed light beam reaches different photocells 44. Control unit 18 responds to the signal from which ever individual photocell 44 is receiving the axially directed beam 43, and translates that photocell’s identity into axial-position information for reference by the user. (PX-1, Col. 3)
45. While the preferred embodiment of the Barnett gyroscope discloses the use of just one light beam projected axially from the rotor, one skilled in the art should recognize quickly, and without undue experimentation, that two light beams would be necessary to resolve the problem of determining the difference between (1) when the rotor merely translated with respect to the stator, and (2) when the rotor’s spin axis axially rotated or precessed with respect to the stator. (Moore Tr. 736-737)
46. For fine resolution of angular pick-off of the rotor’s spin axis, in the preferred embodiment of the Barnett patent, it would be necessary to have not only very small receptor elements at the stator but equally important, one or more very fine light beams projected axially from the rotor. (PX-10; Moore Tr. 736-737)
The Rockwell ESG Gyroscope
47. The accused devices are known as the Electrostatically Supported Gyro Monitor (“ESGM”), which is used in the inertial navigation system of ballistic missile submarines, and the Electrostatically Suspended Gyro Navigator (“ESGN”) AN/WSN-3(V)2, which is used in the inertial navigation system of nuclear attack submarines. (Moore Tr. 422-424) These gyroscopes are manufactured for the Navy, by the Auto-netics Marine Systems Division of Rockwell International at Anaheim, California. (Moore Tr. 373)
48. The predecessor of today’s Rockwell International, has long been involved in the development of precision gyroscopes and inertial navigation systems. Shortly after World War II in 1948, North American Aviation was one of several defense contractors which received technology from the German missile program, and began development of a line of missiles, as well as the inertial navigation systems for such missiles. North American Aviation merged with Rockwell Corporation in 1967, to form North American Rockwell, and eventually, Rockwell International. Rockwell has provided inertial navigation systems for the Minuteman intercontinental ballistic missile, as well as for all Polaris, Poseidon and Trident fleet ballistic missile submarines. (Moore Tr. 369-372, 408-409)
49. Rockwell began to investigate elec-trostatically supported gyroscopes (“ESG”) in 1959. (Moore Tr. 411) In 1968, a miniature ESG was constructed for use in airborne applications. (Moore Tr. 415) This led to the development in 1971, of an ESG for submarine applications. Following competition with a Honeywell ESG, the Rockwell ESG was selected by the Navy for installation on ballistic missile and attack submarines, in 1974 and 1975, respectively. (Moore Tr. 416)
The present form of the accused gyroscopes, also called the micro-ESG, was conceived in 1966, and first built in 1968. (Moore Tr. 412) It contains a solid beryllium rotor one centimeter in diameter. (Moore Tr. 411; DX-31) In principle, however, an ESG rotor needs only to have an electrically conducting surface, and may have an interior made of electrically insulating material. (Allen Tr. 128-129)
*69650. While the ESGM and ESGN systems vary somewhat in physical arrangement and electrical circuitry, both use two electrostatically supported gyroscopes to stabilize the inertial platform on which the accelerometers are mounted. (Moore Tr. 431-433; DX-31, pp. 2-2-4 and 2-2-6; DX-45) That which is most relevant to this litigation, the gyroscope spinner assembly, which consists of the gyroscope rotor and its housing, is identical in the ESG’s used in the two systems. (Moore Tr. 430) A replica of a demonstrative exhibit used at trial to explain the structure of the Rockwell gyroscope spinner assembly, was submitted for the record as DX-46A and 46B. (Allen Tr. 188-194)
51. The rotor of the Rockwell ESG is a solid beryllium sphere, one centimeter (about one-half inch) in diameter. (DX-31, p. 2-2-7) The sphere is embedded with two tantalum wires, set off from the center of the rotor, as shown in U.S. Patent No. 3,880,606. (Moore Tr. 425-426; DX-27) The rotor is surrounded by four pairs of diametrically opposed octantal electrodes mounted on the upper and lower hemispheres of the envelope. The width of the gap between the rotor and the electrodes is approximately 300 millionths of an inch (300 microinches). (Allen Tr. 193) The oc-tantal electrodes are connected to four electrical suspension circuits, which supply a 20,000 Hertz (20 kHz) 300 volt alternating current to the electrodes. (Moore TR. 458-459; DX-28; Fig. 4; DX-31, p. 2-2-7 and Fig. 2-2-9) It produces an electrostatic field strength of approximately one million volts per inch within the gap, which causes the spherical rotor to levitate to the center of the cavity. (Moore Tr. 459-463; DX-31, p. 2-2-9 and Fig. 2-2-13) A set of induction torquing coils surrounds the rotor assembly and produces a rotating electromagnetic field, to initially orient the rotor and bring it to its operational spin speed at 3600 rps (revolutions per second). (Moore Tr. 494-496; DX-31, Fig. 2-2-8) Once operational spin speed is achieved, the induction torquing coils are deenergized, and the rotor continues to spin in the evacuated assembly. The induction torquing coils are not energized when the gyroscope is operating in the normal navigation mode. (Moore Tr. 496-497)
52. The rotor interacts with the octantal electrodes in the manner of a capacitive bridge. Each electrode and the adjacent surface of the rotor, forms a capacitor, in which electrical energy is stored in the electrostatic suspension field between the electrode and the rotor. If forces acting on the device cause the translational position of the rotor to stray from the center of the stator cavity, the capacitive bridge becomes unbalanced, and the external suspension circuitry applies corrective voltages to the electrodes, to return the rotor to the desired position in the center of the stator. (Moore Tr. 473-475; Allen Tr. 210-215; DX-31, p. 2-2-9)
53. A perfectly uniform sphere will rotate so that its geometric center coincides with its center of mass. Since such perfection is impossible to obtain in practice, there will always be some displacement of one from the other. Rockwell decided to intentionally introduce a known amount of mass unbalanced at a preselected position by adding tantalum, to an otherwise solid beryllium rotor. As a result, the ESG rotor does not spin about its geometric center. (DX-31; Allen Tr. 185, 186-187, 195-196) The consequence is that the geometric center’s path describes a plane normal to the spin axis, thus providing a plane of reference perpendicular to, and indicative of, the orientation of the spin-axis of the rotor. (DX-31; Tr. 186-187) In use, the rotor’s surface sweeps out a torodial space, and alternately approaches and recedes from various points on the stator at its rotational frequency, approximately 3.6 kHz, i.e., 3600 cycles per second. (Allen Tr. 199-200; DX-28)
54. The ESG rotor is a solid beryllium sphere, 1 cm. in diameter, to which a small amount of tantalum is added away from the rotor’s geometric center. (DX-31) The added tantalum in early ESG models was “sputtered” onto one-half of the rotor’s exterior, and in later models was imbedded as two or three fine wires inside the *697rotor itself by means of a Rockwell patented process. (DX-27, 28, 29, 30, 31; Moore Tr. 576) Rotors with externally “sputtered-on” tantalum were not truly spherical, i.e., they has a “bump” on one side, and have not been used successfully. (Moore Tr. 532, 573, 748-750)
55. Rockwell rotors with embedded tantalum wires are carefully machined to a spherical shape at about normal room temperatures. (DX-31) At the higher operating temperatures, with a high rotational speed, there is some balancing out of the anisotropic tendency of the beryllium to “prolate” and the centrifugal effect to “oblate ” it, thus causing it to be spherical in use. (Moore Tr. 571; DX-27) This sphericity in use, is vitally important since the rotor must rotate about its principal axis for stability, and the electrostatic forces between the stator charges and the induced charges on the rotor surface, are always normal to the rotor surface. (DX-27; Moore Tr. 574, Allen Tr. 214-216, 222)
56. Since the center of mass of the rotor is slightly offset from its geometric center as a result of the tantalum wires, the rotor rotates about its center of mass, rather than the geometric center, and the surface “wobbles” very slightly at the frequency of rotation. (Moore Tr. 596-605; DX-31, p. 2-2-8 and Figs. 2-2-10, 2-2-11) The amount of wobble is less than 25 mi-croinches, i.e., less than one-tenth of the width of the very thin rotor-stator gap. This is smaller than the thickness of a sheet of paper, and could not be discerned by the unaided human eye. (Moore Tr. 605-610) The electrical effect of the very slight wobble is to redistribute the energy stored in the capacitive electrostatic suspension filed and thus to modulate the 20 kHz suspension voltage at the 3600 Hz rotor spin frequency. This anisotropic phenomenon is identified as mass unbalance modulation (“MUM”). The magnitude and phase of the MUM signal on each of the four electrode plate pairs, is indicative of the angular relationship between the spin axis of the rotor, and a line drawn through the plate pair. External circuitry compares the magnitude and phase relationship of the MUM signals from different plate pairs, and thereby determines the orientation of the rotor spin-axis with respect to the stator. (Allen Tr. 199-202; Moore Tr. 610-612; DX-31, p. 2-2-8 and Fig. 2-2-11) This information constitutes the rotor-position pickoff signal, which is then used to drive the inertial platform’s gimbal torque motors, to maintain the inertial platform in a stable relationship with respect to inertial space. (DX-31, pp. 2-2-5, 2-2-6; DX-45) The rotor-position pickoff concept is described further in Rockwell’s U.S. Patent No. 3,847,026. (Moore Tr. 427-430; DX-30)
57. As previously noted, the rotor is brought to operating speed by induction torquing coils, which are deenergized during normal operation. (DRF-21) Since the stator envelope is evacuated to a very high vacuum, and the spinner assembly is shielded from the earth’s magnetic field by a mumetal enclosure; the rotor can therefore spin for many days before it would coast to a stop, due to the combined friction of the remaining gas molecules, and the electromagnetic braking effect from the residual magnetism which penetrates the mumetal shield. However, since the electrostatic force vector which levitates the rotor is directed toward the geometric center of the ball, a moment arm exists between the force, and the offset mass center of the ball about which it rotates. This provides a small accelerating torque to the rotor, which offsets the decelerating torque, attributable to gas and magnetic field drag. The suspension circuitry is provided with a speed controller, which creates a sharply tuned phase lag between the MUM pickoff signal and the electrostatic force applied to the rotor. The tuned circuit causes the rotation of the ball to be maintained at the desired speed. This concept is described in Rockwell’s U.S. Patent No. 3,906,804. (Moore Tr. 426, 497-499; DX-28)
58. (a) A “capacitor ” is any device that holds an electric charge in response to a voltage ap-plied across two electrically conducting elements called “electrodes”. These electrodes must be separated by an *698electrically non-conducting material called an insulator or a “dielectric”, e.g., either air or, preferably, a vacuum. (Allen Tr. 210-211)
(b) The electric charge held by the capacitor is directly proportional to the capacitance times the voltage, i.e., Q = C X V. In terms of units, one “coulomb ” of charge is the amount held in a capacitor of one “farad ” when one “volt ” of voltage is applied across the capacitor electrodes. (Allen Tr. 211-213)
(c) Given a certain, fixed voltage difference between the two conducting elements of a capacitor, if one moves them closer, one increases the capacitance and hence the charge carried, and vice versa. (Allen Tr. 213)
(d) If one were to apply a source of direct current, e.g., a battery across the electrodes of a capacitor, there would be a build-up of positive charge on the electrode connected to the battery’s positive terminal, and a corresponding equal but negative charge on the other electrode. (DX-23; Allen Tr. 312) While the charge transfer was taking place, there would be an actual flow of electric charge through the wires connecting the battery and the capacitor, i.e., an “actual current ” would flow in the external circuit and light up a light bulb, if one were included in the current path, until the capacitor was fully charged. (Moore Tr. 580-581; DX-44; Moore Tr. 585-586) Therefore, no more actual current can flow through the external circuit when the capacitor is fully charged.
(e) While an actual current was flowing through the external circuit in the preceding illustration, no electrons physically crossed the dielectric and yet, at the end, there is a definite charge separation across the capacitor plates (i.e., the two conducting elements) of the capacitor. (DX-43; Allen Tr. 312) Thus, a capacitor, when it has an externally imposed voltage on it, stores up electrical energy. (DX-40) The charge separation across the capacitor occurs due to the flow of an “apparent current” which does not involve the actual flow of charged particles across the dielectric. The charges merely appear to flow across and out of the capacitor. (Moore Tr. 581-583)
(f) If an alternating current were used in the preceding illustration, i.e., a current varying cyclically in time, so long as there was an actual current in the external circuit, there would also be an apparent current of the same frequency across the capacitor. The capacitor stores up and discharges electrical charge in response to the externally imposed alternating voltage. (Moore Tr. 587)
(g) Such a capacitance of magnitude C, carrying a charge Q, possesses electrostatic capacitive “energy” in the amount of Q2/2C, and can do this much “work ” in the process of being discharged. This capacitive energy is considered to reside in the electric field between the plates, although in fact, the charges of opposite signs, whose separation constitutes the stored energy, actually reside at the electrodes and not in the dielectric between them. Thus, a capacitor, when it has an external voltage applied across it, experiences a separation of charge across it by way of an apparent current, and thereby stores up electrical energy. (DX-40) This is true, whether the applied voltage is constant as in subpara-graph (d), or time-varying as in subpara-graph (f), supra. (Moore Tr. 586-587)
(h) Electrical energy, like any other form of energy, is a conserved quantity, which can be neither created nor destroyed. It may, however, be exchanged among various bodies, or may be converted from one form to another. (DX-40)
(i) The existence of positive charge on one electrode of a capacitor, with an equal amount of negative charge on the other, causes the two elements to be attracted to each other, since opposite charges attract. If one of the electrodes is fixed in space and the other is free to move, then the movable one will try to move towards the fixed one. (Allen Tr. 130, 215-216) Such an attractive force can be used to overcome gravity, i.e., to levitate a body by exerting a force equal and opposite to its weight, or to overcome inertia forces to accelerate a *699free body, per Newton’s First Law of Motion, e.g., to recenter a free rotor in the ESG, when the stator undergoes translational acceleration. (Allen Tr. 130)
59. Immediately around the ESG rotor, and approximately 300 micro-inches larger in radius, is a spherical stator envelope, consisting of two mating hemispheres, plated inside with a smooth electrically conducting surface, which is machined to create eight “octantal stator electrodes ” controlled by an external circuit. (DX-31; Allen Tr. 187-188; DX-45; DX-46a and 46b; Allen Tr. 192-193; Moore Tr. 454, 458-459; DX-28, 29, 30) Despite every attempt by Rockwell to make the eight octantal electrodes identical with each other, “they are not precisely the same”, because “machines have tolerances”. (Moore Tr. 670) Thus, from one stator electrode to another, within a single ESG, and presumably between any two ESGs, there are bound to be differences in individual electrode shape, size, thickness, surface properties and edge smoothness. (Moore Tr. 670-671) In use, therefore, the electrostatic field supporting and centering the rotor, cannot be made truly homogeneous despite specification of very close tolerances. (Moore Tr. 671-672) This small drag due to these stray electrostatic currents on the Rockwell ESG rotor, was never measured by Rockwell, (Moore Tr. 518) nor was this asserted problem addressed by Rockwell in its patent for “Speed Control For An Electrostatically Supported Ball Gyroscope”, apparently indicating that the small drag was not serious enough to impede the proper functioning of the Rockwell gyroscope. (DX-28, Moore Tr. 539)
A variety of factors can induce electrostatic currents in rotors of the Rockwell type. Such factors as inhomogeneity in the stator places, or inhomogeneity in the interior of the rotor, can all introduce small currents in the rotor surface which interact with the stator, and tend to provide slowing-down friction that decelerates the rotor. (Allen Tr. 197)
60. (a) In the ESG, the rotor and the eight stator electrodes are all made of electrically-conducting material separated from the rotor by a strong vacuum, with external voltage differences applied to diametrically opposed stator octants by pairs. (DX-28, 31) The two diametrically opposed electrodes in any one of these four pairs, have charges of opposite sign, i.e., one has positive and the other a negative potential, with respect to the ground for the external circuit connected to both. (DX-31, 28; Moore Tr. 458-459)
(b) At the very start, when the ESG stator is about to be powered-up, there are no voltages applied to any of the stator electrodes, and the spherical rotor surrounded by vacuum, is resting in physical contact with the octant immediately under it. If a voltage difference is now imposed across the pair of opposite octants (on one of which the rotor is resting) a voltage difference is built up between the rotor and the octantal electrode above it, across the vacuum between them. There is now a capacitance in existence, with the upper octantal stator electrode as one of its electrodes, and the rotor surface immediately adjacent to and under it, across the vacuum, as the other electrode. Each carries opposite charges that attract each other. (Allen Tr. 319) This force is proportional to the charge in the plate surface times the electric field between the two surfaces. By a judicious arrangement of capacitors above and below the rotor, and by a circuit that applies the right voltage to these capacitors, it is possible to supply the net force required to hold a rotor vertically suspended, against the force of gravity. (Allen Tr. 130; DX-31; Moore Tr. 591) It is in this manner that the ESG rotor is first levitated. Once it is levitated, it loses physical contact with the stator electrode on which it was initially resting. Therefore, there are now two capacitors, one at the top (between the top stator electrode and the adjacent portion of the rotor surface), and one below (between the stator electrode underneath the rotor and the adjacent portion of the rotor surface), with attractive forces being exerted on the rotor across a 300 micro-inch vacuum gap by each member of the four pair of stator electrodes. *700(DX-45, 31; Allen Tr. 190-194, 215-216) Since these two stator electrodes have charges of opposite signs on them at any instant, the corresponding charges induced on the surface of the rotor adjacent to each, will have charges of signs opposite to these. (DX-43; Allen Tr. 312-313) The rotor, therefore, has no net gain or deficit of electrical charge, and is itself at “ground potential” and represents a “virtual ground”. (PX-4; Allen Tr. 231, 310-314, 214-215)
(c) In the ESG circuit, one externally applied current applied to the four pairs of octantal electrodes, is alternating at 20 kHz, i.e., goes from peak “plus” value to peak “minus” value, with respect to ground, and back, 20,000 times per second. (Allen Tr. 322) The external circuit, through this and other currents at other frequencies, controls the net current to each pair of electrodes, to provide the exact amount of force required at any instant to keep the rotor centered against the effects of both gravity and of any translational accelerations experienced by the vehicle carrying the stator. (Allen Tr. 193-194, 215-216; PX-4; Allen Tr. 231-234) Each electrode pair’s voltage is at a different phase, such that all four acting cooperatively provide the net force needed to keep the rotor centered at all times. A change in gap size will cause an alteration of the currents to octants, and thus activating corrective forces to recenter the rotor. (DX-28; Moore Tr. 458-459, 463, 473-475)
61. Considering only the location and suspension functions of the electrostatic forces exerted by the stator on the rotor, it is obvious that, so far as the free rotor is concerned, there is acting on it a 20 kHz (20,000 cycles per second) basic frequency time-varying voltage from each pair of electrodes. (Allen Tr. 322) The induced charge distribution on the rotor’s surface, therefore undergoes a redistribution of plus and minus charges in a complex pattern at 20 kHz (20,000 cycles per second), in response to these externally imposed voltages, even when the rotor is simply centered at rest in the stator, i.e., when the rotor is not spinning and before consideration is given to how the axial rotor-position pick-off function is accomplished. (Allen Tr. 214-215, 322, 325) When the rotor interior is at a single ground potential, no charges flow through it; all charge redistribution must take place on the rotor by way of surface charges. (Allen Tr. 310-312) In describing a hypothetical gyroscope with a rotor-stator relationship similar to the accused device (DX-43), Dr. Allen hypothesized the existence of electrostat-ically induced charges being present on the rotor, which cause resistance-heat effects as these charges are redistributed on the rotor due to the alternating current nature of the 20 kHz suspension voltage, and the fluctuating rotor-stator capacitance cause by the mass-unbalanced rotation of the rotor. (Allen Tr. 225-226, 320-322)
Dr. Allen testified, however, that the heat created by the circulating currents on the rotor, does not provide information concerning the rotor spin-axis position, and is not directly involved in conveying the pick-off information. (Allen Tr. 326-327) Moreover, there was no testimony from any expert at trial averring that the heat created a problem in the functioning of the accused Rockwell gyroscope. (DX-46A) Absent testimony to the effect that the heat on thé stator presents a sufficient problem in the accused Rockwell device (DX-46A, 31), it cannot be established that the stator in the accused Rockwell device must be cooled down.
It is also presumed that some heat is generated in the Barnett gyroscope pickoff scheme, whereby light energy is converted to electrical energy, and then back into light energy. (DX-42) In regard to this heat energy, however, plaintiff does not assert that this heat energy provides any information regarding the Barnett’s rotor spin-axis position; nor does plaintiff assert that it is involved in conveying any rotor-position pickoff information.
62. In actual use, once the rotor is levitated, it must be spun up to its operational speed. In the ESG, this is done by applying a rotating magnetic field, that first *701induces eddy currents in the rotor body, and, simultaneously interacts with the elec-tro-magnetic field generated by the rotor eddy currents, to provide an accelerating torque to spin up the rotor from rest to a typical operating speed of 216,000 r.p.m. (Allen Tr. 258, 260-261; Moore Tr. 494-496, 497; DX-31, 46B) The rotating magnetic field “motor” is then turned off, and a part of the external circuit, called the “servo circuit", takes control of the rotor’s motion, and either speeds it up or slows it down to the desired speed, and then maintains it at the proper value by continuously supplying energy, applying an input torque to overcome assorted drag torques that otherwise would slow down the rotor. (DX-28; Moore Tr. 497-499, 537-539, 503; PX-3; Allen Tr. 224-226)
63. The rotor is now levitated and, under the combined action of the bias voltage of 20 kHz and the servo voltage of 3.6 kHz, is spinning in the mass unbalanced mode, i.e., the rotation is such that the rotor’s geometric center is rotating about its spin axis. (Moore Tr. 497; DX-31) The surface of the rotor now approaches and recedes from the stator electrodes arrayed about the plane of rotation of the rotor’s geometric center at 3.6 kHz, thereby changing the cavity gap at any point at this cyclic rate. (DX-31; Moore Tr. 596-597) The voltage and charge on each of the rotor suspension plates is, therefore, affected at this 3.6 kHz cyclic rate across the capacitive-bridge-type circuit, and is the origin of the MUM signals used throughout the ESG. (DX-31; Allen Tr. 201-202; Moore Tr. 597-598, 610-611) Whatever the magnitude of this cyclic net voltage is at any electrode, it does not change until there is translational acceleration of the stator, or until the relative angular orientation of the rotor changes with respect to the stator. Any such change can be utilized to center the rotor, or to determine the change in the angular position of the rotor vis-a-vis the stator, i.e., to generate rotor-position pick-off information. (DX-31; Allen Tr. 198-200; Moore Tr. 610-612)
64. The electrostatic force between each octantal stator electrode and the adjacent rotor surface, acts normal to the rotor’s surface. The net stator force on the rotor therefore, acts through the rotor’s geometric center. (Allen Tr. 222) If the rotor were a perfectly balanced sphere then its center of mass (and hence the location of its spin axis) would coincide with its geometric center, and the only consequence of a net external force would be to “center” the rotor. (Allen Tr. 194, 218-219; PX-3; Allen Tr. 222) The ESG rotor, however, creates a situation in which the net force acting through the rotor’s geometric center does not necessarily also pass through the center of mass or the spin axis. (PX-3) There may, in fact, be a short distance serving as a “lever arm" or “moment arm" by which the net force of the stator misses the spin axis and exerts a torque on the rotor. (PX-3; Allen Tr. 220-224, 194-197, 198) This torque can either speed up or slow down the rotor, as may be required, when the electromagnetic spinup field cuts off. (Allen Tr. 198-199; Moore Tr. 497-502; DX-28) Additionally, this torque is used to replenish the losses in kinetic energy of rotation of the rotor by adding energy into the kinetic energy of rotation. (Allen, Tr. 224) No direct tests have been run to determine the drag torques acting on the rotor in the accused Rockwell gyroscope, because such tests were not feasible. (Moore, Tr. 592) However, tests were run by Rockwell on the accused gyroscope, to observe the change in speed of the rotor, or the change in the rotor drift rate, caused by altering factors which set up drag torques. Since the angular momentum of the rotor is known, the magnitude of the drag torques can be calculated, based on the magnitude of the changes in speed or drift rate caused by controlled alteration of the factors causing the drag torques. (Moore, Tr. 592; DX-46A)
65. (a) The ESG rotor’s axial position (stable in the inertial reference frame or moving at a known drift rate), is determined with respect to the moving user vehicle, through an analysis of the voltages at the stator octants by the circuit controlling them. (PX-4; Allen Tr. 230-237; DX-*70231; Moore Tr. 596-599, 610-612) The operation of the suspension and pickoff subsystems in the Rockwell ESG are explained in the Boltinghouse patent entitled, “Cross Product Pickoff For Ball Gyros Of The Electrostatic Levitation Type”, U.S. Patent No. 3,847,026 (“ ‘026”), and is applicable to the Rockwell ESGM and ESGN with respect to Figures 1, 2, 4 and 6a-6d, and the relevant explanation provided in its patent specification. (DX-30; Moore Tr. 427-429) Its suspension electrodes are powered by the 20 kHz bias-current generator 17 which provides a constant bias current I0. (DX-30, Col. 5) When the spherical rotor departs from the center of the cavity, it produces a voltage across transformer winding 23, which is called the rotor-position pickoff signal. The rotor’s axial position is then derived from analysis of the pickoff voltages. (DX-30, Col. 5)
(b) If a strapdown gyroscope is involved, then, as the vehicle turns with respect to the mass unbalanced rotor whose geometric center is rotating in a “reference plane” normal to the spin axis, the relative position of each stator electrode will change with respect to this reference plane. (DX-31) As a result, the closest points of approach of the rotor with respect to each member of any pair of diametrically opposed electrodes, will shift. The proximity of the rotor surface adjacent to each stator octant, determines the charge on each element of the capacitance so formed. (DX-32; Moore Tr. 596-597) This charge on each octant must be modulated by the servo current, to ensure that the precise amount of force is exerted by that stator octant to maintain the rotor in its correct centered position. The angular orientation of the rotor vis-a-vis the stator, is determined from a comparison of the phase relationships of the various mass unbalanced modulation (MUM) pickoff signals, from the four electrode plate pairs. (DX-30; Figs. 4, 6, and accompanying text; Moore Tr. 610)
(c) In the alternative, if the gyroscope is being used in the gimballed mode, the alterations in servo currents to the stator electrodes serve as error signals to control the servomotors at each gimbal bearing. The servomotors act to eliminate these error signals by reorienting the stator so that it is again in congruence with the rotor in the latter’s new position relative to the user vehicle. (Moore Tr. 422) Knowledge of the amount and direction in which the gimballed stator moved with respect to the user vehicle constitutes the rotor-position pickoff information for navigation purposes. This may be used to stabilize a platform carrying three mutually orthogonal accelerometers to determine the user vehicle’s instantaneous translation acceleration, in the inertial reference frame. (Moore Tr. 421-422; DX-31, 45)
66. (a) Given two conducting elements forming the electrodes of a capacitor, i.e., one octant of the stator and the rotor acting as the extension of the opposing octant of the stator, any relative motion between them will affect both the amount of charge held on each electrode and also the amount of energy, i.e., capacitance, stored in the capacitor, provided the voltage difference across the electrodes is maintained constant. (See FF 41) In either case, the force between the two capacitor electrodes will change if the gap changes, while voltage is maintained constant. If, in addition, the voltage difference across the electrode is varying at 20 kHz while the gap is changing the charge of the capacitor at a rate of 3.6 kHz (3600 times per second); then there will be a combined and complex time-dependent variation in the amount of charge at each electrode, in the amount of energy in the eight capacitances at any instant, and in the force between each electrode and the rotor. (Allen Tr. 224-225) This scenario describes, in summary, what takes place between the individual stator octant and the adjacent surface of the rotor in the ESG, while the stator and the rotor are in a fixed angular orientation with respect to each other.
(b) The servo current regulates the amount of charge at each octant in response to the rotor’s position adjacent to it, while corresponding charges of opposite sign move about on the rotor surface, and *703some of the energy is theoretically dissipated into heat, which could possibly increase the drag on the rotor by energizing the residual gas in the rotor-stator gap. (Allen Tr. 321) The same energy source which is providing the servo current, puts this electrical energy back into the electric fields between the various octants, and adds kinetic energy by continually torquing the rotor to maintain its rotational speed and hence the ESG operation. At a given time, an energy flow exists in the accused device between the stator and the rotor. This flow is the electrical energy which flows from one of the eight stator electrodes, which constitutes one electrode of the “eleetrode/rotor and electrode” capacitor, across the rotor-stator gap, to and through the rotor, which is instantaneously functioning as an extension of the opposing electrode/stator octant. (Allen, Tr. 330-333; DX-43)
67. In the Rockwell ESG, it is necessary to replenish the losses in kinetic energy of rotation of the rotor, by providing energy to sustain the kinetic energy of rotation of the rotor. (Allen, Tr. 224) The rotor’s loss of kinetic energy is due to:
(1) residual gas in the rotor-stator cavity; and
(2) induced currents in the rotor which act on stray magnetic fields. (Allen, Tr. 224)
It is impossible, through the Rockwell ESG Mumetal Shield, to shield the Rockwell rotor from all of the Earth’s magnetic forces. (Allen Tr. 224-225) The residual gas, in the Rockwell rotor-stator gap, is heated somewhat by the heat from the Rockwell rotor. The Rockwell rotor apparently generates a small quantity of heat when its surface charges, which are being redistributed about the rotor 3600 times a second, meet the albeit small resistance of the metal of the rotor itself. (Allen Tr. 225-226, 237, 311)
68. If the rotor were a perfectly balanced uniform sphere, its rotation would not modulate the charge distributions and voltages at the stator electrodes. (DX-31) But by having a deliberately unbalanced rotor, i.e., by using the MUM principle, the ESG generates voltage modulation signals whose magnitude and phase, by design, can be related to the reference plane in which the rotor’s geometric center rotates about its spin axis. (DX-27) The reason that the stator electrodes experience a voltage modulation, is that the charge on each stator electrode and its adjacent rotor surface, is related to the geometry of the gap between them, i.e., the gap size and where, say, the center of a particular electrode is located with respect to the rotor’s geometric center. (Allen Tr. 345; DX-31) As the rotor changes its angular orientation vis-a-vis the stator electrodes, its interaction with each individual electrode also changes. Therefore, the contribution of each electrode to the electrostatic field energy changes, and, simultaneously, the physical movement of the rotor repositioning the charges vis-a-vis the different electrodes modulate the voltages at those electrodes. This voltage modulation at the octants, is the information flow that translates in the circuit to axial-pickoff information. (Allen Tr. 343-346) Generating it requires the rotor-held charges to move around in different surface currents on the rotor, using up, through the generation of heat, some of the energy that is carried on the rotor. (Allen Tr. 236-239) For continued use of the ESG, the flow of energy from the stator to the rotor must be maintained, i.e., the rotor’s kinetic energy must be continually replenished, for the user to continue receiving knowledge about its axial position. (Allen Tr. 238-239) However, Dr. Allen was unable to determine how much energy loss occurred. (Allen Tr. 239)
69. There is a storage of electrical energy, i.e., capacitance, in the eight capacitors formed by the octantal electrodes of the stator, the eight rotor surfaces directly adjacent thereto, and the gap between each therein. (Allen Tr. 213-215) The transfer of energy in and out of these capacitors does not constitute a transfer of energy in and out of the rotor itself. Rather, the energy of each capacitor is stored in the gap between each octantal electrode of the stator and its respective adjacent rotor sur*704face. (Allen Tr. 213-215, 311; DX-40) Any change in the physical relationship of one of the eight given capacitors, i.e., one of the single electrode-octants of the stator and its adjacent rotor surfaces, i.e., a change in the relative angular orientation of the rotor, causes either an increase or decrease in the capacitance of the particular capacitor. (Allen Tr. 213) As a particular octant of a rotor moves closer to a particular octantal electrode of the stator, the capacitance, i.e., the electrostatic energy in that rotor-stator gap, increases. (Allen Tr. 213) If that same gap widens, the electrostatic energy in that gap decreases. (Allen Tr. 213; DX-40) It is this change (modulation) of capacitance (electrostatic energy in each rotor-stator gap) within each of the eight rotor-stator gaps (gaps between each single electrode-octants of the stator and its adjacent rotor surface) that provides the information to the computer which then discloses the new physical orientation of the rotor within the stator in the Rockwell ESG (also called the rotor-position pickoff scheme). (Allen Tr. 213-215, 229-237; PX-4; DX-40, p. 414, 2nd col., 3rd para.) During the spinning of the rotor within the stator, at approximately 3600 rpm, there is a rapid change in the electrical charges, on the surface of each stator-electrode octant and its adjacent rotor surface, for each single revolution or spin. (DX-30, 31; Allen Tr. 237-238) This rapid change or redistribution of charges on the respective surfaces of each stator-electrode octant and its adjacent rotor surface, causes a small loss of energy in the form of heat in the rotor and the stator. (Allen Tr. 237) This loss in heat, which both parties agree is very small, has apparently never been measured and, therefore, inferred to be insignificant. (Allen Tr. 238-239)
NON-INFRINGEMENT
70. Claim 7, for element-by-element analysis, may be written as follows:
(a) A control means, comprising:
(b) a stator means
(e) and a movable means;
(d) locating means for locating said movable means in a certain location with respect to said stator means;
(e) first energy means, responsive to movement of said movable means away from said certain location, for energizing said locating means;
(f) second energy means for imparting rotation to said movable means; the said movable means being rotatable in response to energy from a second energy means;
(g) third energy means for imparting rotation to said movable means; the said movable means being rotatable in response to energy from a second energy means;
(h) indicating means carried by the stator means, responsive to energy of said third energy means, to indicate relative movement of the axis of the movable means with respect to the stator means;
71. Claim 7 of the patent in suit is written in the means function language as permitted by 35 U.S.C. § 112, which in pertinent part reads:
An element in a claim for a combination may be expressed as a means for step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof [Emphasis supplied.]
Thus, in order to determine the parameters of claim 7, we must look to the “structures” described in the Barnett specification. When we do so, it is clear that the following structures, which correspond to the following claimed means-for elements, are covered by claim 7:
stator means : stator 12 in figure 1 of Barnett patent; DX-41; DX-1, page 3; Moore Tr. 773;
movable means : rotor 10 in figure 1 of Barnett patent; DX-41; DX-1, page 3, 4; Moore Tr. 773;
*705locating means (for locating said movable means in a certain location with respect to said stator means) light beams 26, photo cells 30, and control unit 18 in figure 1 of DX-41; DX-1, page 5; Moore Tr. 774-76;
first energy means (responsive to movement of said movable means away from said certain location, for energizing said locating means) electromagnets 22 in figure 1 of DX-41; DX-1, page 5; Moore Tr. 776-79;
second energy means (for imparting rotation to said movable means; the said movable means being rotatable in response to energy from a second energy means) light energy 36, from light sources 34 on stator. Some light energy 36 is used to power rotor photocells 46. See Figure 1, DX-41; DX-1, page 6; Moore Tr. 779, 789;
third energy means (carried by the movable means; the said third energy means being actuated by said second energy means) photocells 46 on rotor vanes 38 send current to light bulb 40 in figure 1, DX-41; DX-1, page 6; Moore Tr. 783-88;
indicating means (carried by the stator means, responsive to energy of said third energy means, to indicate relative movement of the axis of the movable means with respect to the stator means) photocells 44 on stat- or with light beam 43 in figure 1, DX-41; DX-1, page 7; Moore Tr. 790.
72. In regard to “stator means”, both the disclosed patented device and the accused Rockwell ESG device contain a “stator means”. (Moore Tr. 773)
73. In regard to element “movable means”, both the disclosed patented device and the accused Rockwell ESG contain a “movable means”, i.e., their respective rotors. (Moore Tr. 773)
74. In regard to “locating means ” and a “first energy means, the patented device and the accused Rockwell ESG device substantially differ. In the disclosed patented device (the disclosed Barnett gyroscope), two distinct subsystems are required for these two elements, i.e., the “locating means ” consisting of light beams 26, photocells 30 and control unit 18; and the “first energy means” consisting of electromagnets 22. In contrast, however, in the accused Rockwell ESG, the funtions of both the “locating means" and the “first energy means” are performed by a single element, i.e., the stator octants. (Moore Tr. 775, particularly lines 17-20; Moore Tr. 778-779, particularly lines 2-5 at Tr. 779)
75. In regard to the “second energy means ”, the disclosed patented device and the accused Rockwell ESG device again substantially differ. In the disclosed patented device, a single subsystem, the light energy 36 that falls on the rotor vanes 38, fulfills the requirement of the “second energy means ”. In contrast, however, in the accused Rockwell device, two separate subsystems, i.e., the electromagnetic coils, and the suspension servo, fulfill the functional requirement of the “second energy means ”. (Moore Tr. 779, 780, particularly lines 6-7 at Tr. 780)
76. In regard to the “third energy means carried by movable means”, the disclosed patented device and the accused Rockwell ESG again substantially differ. In the disclosed patented device (the Barnett gyroscope) lightbulb 40 on rotor 38 fulfills the requirement of “third energy means carried by movable means ”. (Moore Tr. 780) In contrast, however, in the accused Rockwell ESG, no one separate element performs the sole function of the “third energy means carried by the movable means ”. Plaintiff asserts that the accused Rockwell ESG’s rotor, which has already been designated as fulfilling the “movable means”, also fulfills the “third energy means ”. Plaintiff further advances the novel theory that the accused Rockwell ESG’s rotor, performs the additional dual role fulfilling the function of the “third energy means” also. Therefore, assuming, arguendo, that plaintiff is correct, the disclosed patented device and the accused Rockwell device still significantly differ because the accused device does not have a separate subsystem for the “movable means” and the “third energy means ”, while, in contrast, the disclosed patented device contains separate subsystems for the “movable means ” and “third energy means ”. (Moore Tr. 780-787) *706Moreover, while the patentee is his own lexicographer, there is no rationale for using the claim language “energy means carried by the movable means ” unless it was intended to refer to a specific active energy source, as opposed to the “rotor” itself, which is already covered by the “movable means”. (Moore Tr. 781, 783-784, 791-793, particularly lines 19-22 at Tr. 784) The accused Rockwell ESG’s rotor, rather than being an active rotor as asserted by plaintiff, is more properly described as an inactive rotor, or a passive rotor. It is in fact merely a passive element which possesses energy which is impressed upon it. (Moore Tr. 792-794 particularly lines 6-12 at Tr. 794) Finally, plaintiffs attorney admitted indirectly that its literal infringement case was frail, at best. Plaintiff admits that “[t]he structure of the Rockwell ESG, in some respects, is different from that of the Barnett gyroscope disclosed in the illustrative embodiment.” See plaintiffs post-trial brief at 14. Moreover, plaintiffs substantial efforts have been focused upon establishing his infringement argument based on the doctrine of equivalents, rather than on literal infringement. See, e.q., plaintiffs post-trial brief at 14-21.
Doctrine of Equivalency
77. The “spin-up and speed control” structures involved in the Barnett gyroscope vis-a-vis the accused Rockwell gyroscopes, are substantially different. (Allen Tr. 280)
78. Moreover, their principles of operation of the Barnett gyroscope vis-a-vis the accused Rockwell gyroscopes are also substantially different. (Allen Tr. 280-91) On the one hand, the Barnett gyroscope’s drive subsystem is based upon the radiometric principle previously discussed. On the other hand, the accused Rockwell gyroscopes use multiple subsystems that utilize: (1) induction coils for spin-up; (2) the mass unbalanced principle (MUM) with the two off-centered tantalum wires imbedded in the rotor itself; (3) an electrostatic suspension field; and (4) a speed-control circuit hooked up to the suspension field. Thus, the Barnett drive subsystem and the Rockwell drive subsystem are based on principles of operation substantially different from each other. (Allen Tr. 281-283) In addition, the manner of - operation of the Barnett rotor-drive subsystem and the Rockwell rotor-drive subsystem is substantially different. (Allen Tr. 283-284)
79. The suspension subsystem of the Barnett patent, which is based upon the use of an electromagnetic field, and the suspension subsystem utilized in the Rockwell gyroscope, which is based upon the use of an electrostatic field, operate on substantially different principles. (Allen Tr. 287-289) Moreover, their mode of operation is also significantly different. (Allen Tr. 289)
80. The modes of the position-control system of the Barnett device; i.e., the photoelectric system wherein light beams are directed along the sides of its rotor and the degree of blocking by the rotor, if any, is sensed by photocells on the opposite side of the Barnett stator, vis-a-vis, the position control system of the Rockwell gyroscopes, i.e., where the translational shifting of the rotor itself causes an increase of capacitance between the rotor and its adjacent octanal stator electrode on one side and a decrease on the opposite side of the rotor and its adjacent octantal stator electrode, thereby unbalancing the respective capaci-tances, which unbalancing is then sensed, are substantially different. (Allen Tr. 289-91)
Validity
81. Defendant, in asserting invalidity based upon a combination of lack of utility and lack of enablement, first asserts the unworkability of the following aspects of the Barnett gyroscope: (1) translation-rotation pickoff interaction; (2) electromagnetic suspension; (3) radiometer drive; (4) rotor instability; (5) rotor-position pickoff resolution; (6) rotor-stator gap; and (7) control circuitry. Defendant further asserts that, even assuming these problems can be met, they can only be met through the applica*707tion of post-filing-date technology. We disagree.
82. Each and every problem adduced by defendant was straightforwardly solved by plaintiffs, either through their expert witness, Dr. Allen, or on cross-examination of defendant’s expert witness, Mr. Moore. In addition, no substantial evidence was presented by defendant to prove that any of these solutions were necessarily the result of undue experimentation.
83. Defendant’s only witness, Mr. Moore, presented as a person skilled in the art of gyroscope design, testified to an assortment of alleged shortcomings in the illustrative embodiment presented in the Barnett specification. During his cross-examination, however, Mr. Moore, still testifying as one reasonably skilled in the art of gyroscope design, repeatedly conceded either that he nor anyone else at Rockwell had personal knowledge of any of the asserted problems regarding the Barnett design, or whether persons skilled in the art could develop simple solutions to resolve such problems. The evidence adduced at trial makes it clear that no problems with the Barnett invention are beyond the capacity of persons skilled in the art to solve in relatively simple methods and without undue experimentation.
Translation-Rotation Pickoff Interaction
84. Defendant asserts that in the Barnett rotor-position pickoff scheme, with only one axially directed light beam 43, there could be confusion and false readings because photocells 44 would not be able to distinguish between rotor translation and its rotation vis-a-vis the stator.
On cross-examination, however, Mr. Moore conceded that one skilled in the art could easily solve the problem by having two axially directed light beams emitted on opposite sides of the rotor. (PX-10; Moore Tr. 736-39) Moreover, even with only one light beam directed axially, if the gyroscope’s suspension system were prompt in correcting any translational movement of the rotor from its centered position, any confusion would be of very short duration. (Moore Tr. 740-741)
Electromagnetic Suspension
85. Defendant adduces that the use of an electromagnetic field to suspend the rotor, as well as the inhomogeneity of the Barnett rotor will give rise to error torques and variable draft rates.
On cross-examination, however, Mr. Moore stated without hesitation: “I am highly confident that no one at Rockwell has attempted to make the Barnett gyroscope work, even as a study,”; that “No one at Rockwell has attempted to analyze or fabricate the particular embodiment that is shown in the Barnett patent,” and that “Rockwell has not proven that the design would not work, that’s correct.” (Moore Tr. 694-696) Mr. Moore was indulging in conjecture in alleging this problem as proof of the impracticality of the Barnett patent. Moreover, Mr. Moore admitted that the pri- or art Parker patent, which discloses electromagnetic suspension, describes methods for overcoming these precise difficulties, provided a high degree of magnetic homogeneity is obtained in the rotor. (Moore Tr. 466-467; DX-8, Col. 2/6 — Col. 3/36) In response to the court’s questions, Mr. Moore asserted that one skilled in the art could not overcome such problems with the Barnett rotor design. (Moore Tr. 466-467) This should be weighed, however, against his earlier testimony that neither he nor anyone else at Rockwell had even made a study, let alone careful experiments, to see if the Barnett design could work. (Moore Tr. 694-696) Mr. Moore also admitted a prejudice against electromagnetic techniques at Rockwell going back to his early days there. (Moore Tr. 437-438)
Radiometer Drive
86. Defendant maintains that the radiometric technique for spinning the Barnett rotor requires the presence of some gas, and therefore would not generate sufficient torque, hence sufficient speed for stable operation, because of the gas frictional drag.
*708When the court asked Mr. Moore for the basis for his allegation, he replied “It’s based — I suppose one could call it conjecture.” [Emphasis supplied.] (Moore Tr. 479) Mr. Moore also admitted that he had performed no calculations on the subject: (Moore Tr. 479-480)
The Court: But you didn’t prove the unworkability of it through a mathematical formula, is that your testimony?
Mr. Moore: That’s correct.
(Moore Tr. 480)
Moreover, Mr. Moore admitted that he had not conducted even the simplest kind of experiments to confirm his assertions:
The Court: Have you personally done any experiments plotting out the interference of gas in a gyroscope such as the Barnett gyroscope, say with one percent gas, close to a vacuum, half a percent, five percent, 10 percent, 20 percent, and to see if that was a straight-line proportion or a curve relationship or what relationship?
Mr. Moore: Not exactly as you have related the question, I have not, to determine the quantitative relationship between drag and specific amount of gas present.
The Court: So you don’t know exactly at, say one percent of gas in the vacuum or in the partial vacuum how drastic that detrimental effect would be and whether or not that effect would cause the gyroscope to be substantially inefficient rather than just very close to efficient.
Mr. Moore: No.
(Moore Tr. 485-486)
Finally, Dr. Allen, plaintiffs’ expert witness, testified earlier that the radiometric technique would work, even if the gas around the rotor was at atmospheric pressure. (Allen Tr. 145) Moreover, Dr. Allen’s calculations indicate that with hydrogen around the rotor, the Barnett rotor could rotate with more than sufficient speed. Finally, on cross-examination, Mr. Moore conceded that one skilled in the art might consider utilizing that phenomenon by reducing the gas pressure sufficiently, thus reducing frictional drag, and obtaining high torque by means of large vanes. (Moore Tr. 731-733)
Asserted Rotor Instability
87. Defendant avers that the complexity of the construction of the Barnett rotor will cause physical misalignment and differential thermal expansion, as well as spin-axis wobble precluding stable rotation of the rotor. Defendant further asserts there is no provision to damp rotor polhode motion or to torque the rotor to the preferred orientation. Pholode means the motion of the major axis of inertia in seeking the spin axis.
On cross-examination, however, Mr. Moore admitted that one skilled in the art could select rotor materials so as to reduce thermal instability. (Moore Tr. 679-680; 706) Mr. Moore also agreed that the Barnett stator could be turned in an adjustable mounting, so as to be initially aligned with the rotor’s spin axis, no matter how both were then aligned with respect to a chosen inertial reference frame, and that one could then program a computer to compensate for any corrections “for the axis being pointed in a way other than the way you want.” (Moore Tr. 687-693) Moore added, that by following the motion of the Barnett axially directed beam, one could compute and compensate for polhode motion as well. (Moore Tr. 700-702) This testimony reflected the use of the Barnett gyroscope in the strapdown mode. (Moore Tr. 688)
Mr. Moore further conceded that speed control is unnecessary, provided the rotor’s speed is known. (Moore Tr. 486) On cross-examination, he admitted that the Nordsieck patent taught, as early as 1961, that the precession speed of a gyroscope rotor is related to its rotational speed; that it can be extracted from the axial rotor-position pickoff signal; that the Barnett rotor would precess; and that, therefore, per Nordsieck, the Barnett rotor speed could be obtained. (Moore Tr. 696-702) Mr. Moore further stated that one reasonably skilled in the art would consider such a technique to solve the problem he had identified. (Moore Tr. 703) Moreover, he stat*709ed that each user would have to select a rotor speed in light of the particular application he had in mind. (Moore Tr. 685-687)
Finally, it should be noted that Mr. Moore was talking about only the illustrative embodiment, and that he was careful to say only that “one would anticipate a great deal of difficulty in assembling all of these different components in a precisely aligned fashion.” (Emphasis supplied.) (Moore Tr. 490) On cross-examination, Mr. Moore confessed that in principle, while difficult, it would be possible for one skilled in the art to balance even a complex rotor:
Mr. Nirmel: Isn’t it correct that you yourself have no personal knowledge and that you are merely conjecturing that there would be a significant unbalance of the rotor that would affect the Barnett gyro operation?
Mr. Moore: That’s correct. I was simply referring, I believe, to the apparent difficulty in achieving an adequate degree of mass balance with such a complex assembly.
Mr. Nirmel: But in principle it is possible to devise techniques to balance complex rotors, is it not?
Mr. Moore: In principle, it is. In practice it has proven more difficult, the more complex the assembly.
(Moore Tr. 703)
Rotor-Position Pickoff Resolution
88. Defendant claims that the large number of components to be carried by the interior surface of the stator, and the concept of an optical photocell rotor-positions pickoff array is impractical and cannot provide sufficient rotor-position pickoff angular resolution to construct a useful gyroscope.
Assuming, arguendo, the truth of the matter asserted, the law is clear that a patented device does not have to compete, on the basis of accuracy, resolution, etc., with later developed more sophisticated allegedly infringing devices. See Decca Limited v. United States, 544 F.2d 1070, 1077, 1080-1081, 210 Ct.Cl. 546, 558, 564-565 (1976). Demonstrating that the 1963 Barnett application may not have met the needs of today’s rotor-position pickoff resolutions, e.g., to the VmooÜi of a degree, or even Vsoth of a degree, does not prove the per se unworkability of the Barnett gyroscope.
Rotor-Stator Gap
89. Defendant argues that the multiple independent optical systems used in the Barnett light beam gyroscope, requires a large rotor-to-stator gap which is contrary to normal gyroscope design practice.
On cross-examination, however, Mr. Moore conceded that the Barnett disclosure does not limit the number of locator beams 26, and that by increasing the number of such beams one could reduce the gap provided one could locate the additional lights 28 and photocells 30 at the stator. (Moore Tr. 767-770) He also conceded that one skilled in the art might consider making the stator of a transparent material, using fine laser beams, and locating paraphernalia like lights 28 and photocells 30 on the outside surface as a way to effectuate this simple solution to reduce the gap size. (Moore Tr. 770-771) Moreover, both the Parker and Nordsieck patents suggest a glass stator that will transmit light and not affect the magnetic suspension field. (DX-8; DX-20)
Finally, while Mr. Moore was testifying as an expert in gyroscope design in generally, it is undisputed that Mr. Moore had never, in his professional career, worked on or with electromagnetically supported rotors in gyroscopes.
Control Circuitry
90. Defendant asserts that there is no control circuitry shown for what is a very complex control situation due to the necessity for interaction among the multiple independent optical systems in the Barnett gyroscope.
While there is no detailed control circuitry disclosed in the Barnett specifications, it should be noted that neither is any de*710tailed control circuitry claimed. The new, useful, and nonobvious teaching of the Barnett invention resides in the active, spherical, free rotor — not in minute details of control circuitry. Defendant did not carry its burden of proof insofar as to demonstrate that such required circuitry was not within the scope of one skilled in the art in 1963.
As a final argument for invalidity, based on a combination of Sections 112 and 101, defendant claims that plaintiffs’ adduced solutions, supra, are not relevant because they are premised on today’s technology, rather than that of 1963. We disagree.
Defendant’s assertions are not persuasive enough to sustain its burden of proof in establishing a prima facie case for invalidity, in view of the presumption of validity the Barnett patent enjoys, particularly because defendant’s sole expert was unsure and hesitant in his responses, even on direct testimony to relevant questions of his own counsel, e.g.,:
Mr. Daigle: Was the state of fiber optic technology in April 1963 sufficient to achieve that solution, if you know?
Mr. Moore: I do not know, but I do not believe it was.
(Emphasis supplied.) (Moore Tr. 797)
We do not believe, relying on the wavering type of testimony seen above, that we could find that all the technologies suggested by plaintiffs to resolve the apparent problems asserted by defendant, were post-filling-date in character. We hold, therefore, that with regard to sections 112 and 101, that defendant has not carried its burden of proof as imposed by section 282.
Finally, even if some of defendant’s arguments of unworkability of some aspects of the Barnett gyroscope were true, because they allegedly utilized post-filling-date technology, the defense of non-utility cannot be sustained without proof of total incapacity. E.I. duPont de Nemours & Co. v. Berkley and Co., 620 F.2d 1247, 1260, fn. 17, 205 U.S.P.Q. 1, 10 (8th Cir.1980). Moreover, just some degree of utility is sufficient for patentability. Id. Finally, the fact that an invention has only limited utility and is only operable in certain applications, is not grounds for finding lack of utility. Raytheon Co. v. Roper Corp., 724 F.2d 951, 958-959, 220 U.S.P.Q. 592, 598 (Fed.Cir.1983); Carpet Seaming Tape Licensing Corp. v. Best Seam, Inc., 694 F.2d 570, 578, 216 U.S.P.Q. 873, 880 (8th Cir.1982).
Unobviousness
91. While defendant did not specifically address the issue of obviousness, nevertheless, it asserted invalidity, albeit premised on section 112 rather than section 103. Since obviousness is subsumed under validity, and pursuant to the instruction from the Federal Circuit, that the lower court “should decide validity and infringement and should enter a judgment on both issues when both are raised in the same proceeding,” out of an abundance of caution, we also address unobviousness.
The applicable statute is, of course, 35 U.S.C. § 103:
A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negatived by the manner in which the invention was made. [Emphasis supplied.]
92. At trial, however, defendant did not present any evidence to the court to rebut the presumption of validity, as it applies to the issue of obviousness. Defendant did: (1) not present any prior art more relevant that that before the Patent Office; (2) not establish a prima facie case of obviousness; (3) not present testimony by its sole expert on the issue of obviousness, i.e., the four-part obviousness analysis discussed, supra; (4) not address the obviousness issue in its post-trial brief regarding liability issues; and (5) not propose any findings of *711fact regarding obviousness. It is therefore clear, a priori, that defendant has not carried its burden, i.e., adducing clear and convincing proof that the claimed Barnett invention, as a whole, as described by claim 7, would have been obvious to one skilled in the art in 1968, the filing date of the Barnett patent application. Because of this void in the evidence of record regarding the obviousness issue, this court need not further address the issues subsumed under obviousness. We therefore find that the defendant did not carry its burden of establishing invalidity, through obviousness, of claim 7 in suit, as required by 35 U.S.C. § 282.
CONCLUSIONS OF LAW
Upon the foregoing findings of fact, which are made a part of the judgment herein, the court concludes as a matter of law that claim 7 of the patent in suit is valid, but not infringed by the Rockwell ESG devices. Plaintiff, accordingly, is not entitled to recover, and its petition is therefore to be DISMISSED.
. Assuming, also, no prosecution history estop-pel is applicable.
. There is presumably some heat generation, also, in the Barnett gyroscope pickoff scheme, *653whereby light energy is converted to electrical energy, and then back into light energy. But this “heat", as in the Rockwell ESG, does not provide information concerning the rotor spin-axis position, and is also not directly involved in conveying the pickoff information.
. Hammond, U.S. Patent No. 1,467,154; Witt-kuhns, U.S. Patent No. 1,999,646; Ammen, U.S. Patent No. 2,378,744; and Dias, U.S. Patent No. 2,541,217.
. The techniques known in the art, in the relevant years, fall into two categories characterized *667by whether the rotor (a) plays merely a passive role, or (b) actively participates, i.e., receives energy, converts it to a different form of energy, converts it back to the original type of energy, and sends this energy in cooperation with means carried on the stator, for use in the process which provides the user with accurate rotor spin-axis position information on a continuous basis for prolonged periods.
. Defendant’s Post Trial Brief, Vol. I, p. 29.
. Post-filing-date state of the art cannot be used to construe claim language. See W.L. Gore & *674Associates, Inc. v. Garlock, Inc., 721 F.2d 1540, 220 U.S.P.Q. 303 (Fed.Cir.1983).
. Court of Customs and Patent Appeals.
. 35 U.S.C. § 112 (1976). See In re Barker, 559 F.2d 588, 593, 194 U.S.P.Q. 470, 474 (CCPA 1977), cert. denied, 434 U.S. 1064, 98 S.Ct. 1238, 55 L.Ed.2d 764, 197 U.S.P.Q. 271 (1978), where the CCPA carefully distinguished between the written description and enablement requirements of § 112, first paragraph.
. In re Moore, 439 F.2d 1232, 1236, 58 CCPA 1042, 169 U.S.P.Q. 236, 239 (1971) (Emphasis supplied).
. Section 101 reads:
Inventions Patentable
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
. Linde Air Products Co. v. Graver Tank & Mfg. Co., 86 F.Supp. 191, 197, 75 U.S.P.Q. 231, 236 (N.D.Ind.1947), rev’d, 167 F.2d 531, 536-537, 77 U.S.P.Q. 207, 212 (7th Cir.1948), aff’d, Graver Mfg. Co. v. Linde Air Co., 336 U.S. 271, 277-279, 69 S.Ct. 535, 538-39, 93 L.Ed. 672 (1949).
. General Electric Co. v. United States, 206 U.S. P.Q. 260 (Ct.Cl., Trial Div.1979), aff’d, 654 F.2d 55, 211 U.S.P.Q. 867, 228 Ct.Cl. 192 (1981) (Emphasis supplied).
. Moore Tr. 479-80.
. Moore Tr. 485-86.
. Polhode means the motion of the major axis of inertia in seeking the spin axis.
. Moore Tr. 687-693.
. Moore Tr. 703.
. Moore Tr. 797.
. Lindemann, 730 F.2d at 1463.
. Pursuant to Sernaker and Environmental Designs, it could be argued that the Graham analysis has now been amended to create a four-prong approach, rather than a three-prong approach, to the question of obviousness.
. Lindemann, 730 F.2d at 1459.
. The techniques known in the art, in the relevant years, fall into two categories characterized by whether the rotor (a) plays merely a passive role, or (b) actively participated, i.e., receive, convert, and utilize the resultant energy, in cooperation with means carried on the stator, in the process which provides the user with accurate rotor spin-axis position information on a continuous basis for prolonged periods.