dissenting.
The district court did not follow this court’s precedent in deciding the question of infringement. Nor does the majority, which has, in addition, undertaken to re*1536write this precedent. Thus I respectfully dissent.
Questions of patent infringement turn on the application of a legal document, in the form of patent claims, to technology, as embodied in an accused device. Perkin-Elmer and Westinghouse both presented extensive evidence on the technology. The testimony of opposing technical experts was refreshingly harmonious; perhaps because these principles of classical electricity are beyond debate. The majority has, curiously, created substantive technological differences where there are none, adopting an incorrect view of the invention and a misapplication of the law.
Perkin-Elmer asserted, and reaffirmed in response to a question from the bench, that its position was not that it had made a two-element combination such that all other claim limitations should be ignored. Undoubtedly a basic contribution of the invention is the use of a helical resonator to ignite and power an electrodeless discharge lamp, but Perkin-Elmer has never argued that the other limitations in the claims are irrelevant. To the contrary, Perkin-Elmer has argued that the various claim limitations are part of its claimed invention and appear in the accused devices. Perkin-Elmer has not asked the court to “erase a plethora of meaningful structural and functional limitations”, as the majority states. The emphasis on Perkin-Elmer’s “two-limitation claim”, and significant portions of the technological discussion, are creations of the majority, not Perkin-Elmer.
The majority has found non-equivalence between the known transformer used by Westinghouse and the known transformer of the '884 claims, although the references of record teach both with helical resonators. Unless the doctrine of equivalents is to be eliminated as an equitable tool of infringement analysis, these simple electrical equivalents are squarely within it. It is not useful to scientists and engineers for courts to magnify scientific differences beyond their reasonable meaning.
If indeed the majority wishes to redefine this court’s precedent on the doctrine of equivalents, as in the majority’s footnote 8, this important issue should be confronted directly, as a matter of public policy. The patent-user community is entitled to rely on judicial decisions as they exist and until they are overruled as a matter of law, not as some might prefer to have written them.
I
The devices at issue contain multiple elements, and the claims contain many limitations. Perkin-Elmer has not claimed a two-element invention, and has not sought enforcement in those terms; as will be apparent from a review of the devices and the claims at issue.
The Devices
The electrodeless discharge lamp (EDL) was a known spectral line light source. At the time of the invention of the ’884 device EDL systems were of limited practical use as light sources in atomic absorption spectroscopic analysis, due to their high cost, large size, unreliable operation, and undesirable electromagnetic radiation. However, EDLs had the capability of analysis of certain chemical elements not readily analyzable by the hollow cathode lamp then in use. Both Perkin-Elmer and Westinghouse recognized a need for EDLs having the reliability, low cost, and small size that would enable them to be routinely used for chemical analysis in existing spectrophotometric equipment.
In 1970 Westinghouse started a research program on EDLs, assigning engineers Robert M. Hruda and George K. Yamasaki to investigate lamps operating at radio frequencies below 915 megahertz, and Walter Hayter to investigate EDLs operating at microwave frequency of 2450 megahertz. Hruda, who was described as a distinguished engineer and who testified as a Westinghouse expert witness, and Yamasaki learned from the technical literature that EDLs operating at microwave frequencies were easier to start and had a longer expected life, and in the course of a two-year research program developed a 915 megahertz microwave cavity/EDL combination that met the requisite size limitations. Westinghouse exhibited this device *1537at a national conference on Analytical Chemistry and Applied Spectroscopy in March of 1973. Hruda was granted U.S. Patent No. 3,826,950 for this microwave EDL, but the device had several inadequacies and was not commercialized.
In 1971 Perkin-Elmer was also trying to develop an EDL for the same use. PerkinElmer’s engineers selected the radio frequency of 27.12 megahertz, for reasons that were explained at trial, but the devices they developed required a separate ignition coil, were not self-starting, and the impedance match was not good. In 1972 Fred C. Gabriel was asked to try to solve these problems.
Gabriel also selected the radio frequency of 27.12 megahertz. But he created an entirely different EDL system using a helical resonator, a system not previously used with EDLs. His system eliminated complex starting circuits and was self-starting if the lamp went out during use. It met all the goals of reliability, size, economy, and effectiveness. This is the device of the ’884 patent.
The Gabriel device was exhibited at the same March 1973 conference on Analytical Chemistry and Applied Spectroscopy. As found by the district court, Westinghouse sought to obtain the Perkin-Elmer device in order to learn the details of its construction. Perkin-Elmer Corp. v. Westinghouse Electric Corp., No. 77 C 1923, slip op. at 16 (E.D.N.Y. Aug. 19, 1985). Meanwhile Westinghouse began development of a device operating at the radio frequency of 27.12 megahertz, which it had learned at the exhibit. In early June 1973 Hruda designed circuitry that met the dimensional requirements Perkin-Elmer had achieved, but his device was considered unsuitable because of its sensitivity to temperature variations, and was subsequently abandoned.
The district court found that it was not until July of 1973, when the Perkin-Elmer device was obtained and dismantled, that Westinghouse learned that the Perkin-Elmer device had a quarter-wavelength helical resonator coil, and that the coil was open circuited at one end, and coupled into the r-f power supply by a transformer. Id. at 16-17. By September of 1973 Hruda had constructed a 27.12 megahertz EDL using a quarter-wavelength helical resonator coil, open circuited at one end, and coupled into the r-f power supply by a transformer. This device was commercialized. It contained all the elements of the Perkin-Elmer device, but used a known alternative form of transformer. It was called Westinghouse I at trial.
Westinghouse later developed Westinghouse II, identical to Westinghouse I except that some of the turns of the quarter-wavelength coil were replaced with a capacitor. The district court stated: “The alterations were known prior art options for helical resonator circuits.” Id. at 18.
The Patented Invention
Gabriel’s ’884 patent, entitled “Electrode-less Discharge Lamp and Power Coupler Therefor”, describes the purpose of the invention as follows:
to provide a spectral source electrodeless discharge lamp operating unit which includes an inductive resonator coupler, for coupling r-f electric power into the lamp for starting and operating it, that provides efficient coupling, that is simple and economic to construct and operate, that provides the requisite voltage and current ratios for starting the lamp and then for maintaining it in operation without intermediate switching means or alternative power supply connections, and that has an automatic self starting capability if the lamp goes out during operation.
Gabriel’s invention is claimed in the standard way, as the combination of its several components; see claim 1 in the majority opinion. Although the majority has ignored all of the claim limitations except those of claim clauses (h) and (i), which relate to the transformer coupling to the r-f power source, Gabriel’s invention is the entire combination, as illustrated in the patent drawing:
*1538[[Image here]]
The inductive resonant coupler (11) provides both the high voltage electrostatic field needed to start the EDL (10) and the electromagnetic field needed to run the EDL by r-f power from generator (12). The resonant coupler is a cylindrical shield (14) made of conductive material, with a helical coil (15) concentrically spaced inside the cylinder.* The coil is of a length equal to one fourth the wavelength of the resonant frequency of the r-f power. One end of the coil is grounded at (18); the other end is open circuited, except for the optional provision of a trimming capacitor (32) between the cylindrical shield and the helical coil, as a fine tune adjustment of the resonant frequency. The coil is coupled to the r-f generator at tap (26) by connector (25) and coaxial cable (27).
At the frequency of the Gabriel design, 27.12 megahertz, the helical resonator is capable of a substantially higher electrical “Q” value than the prior art tank circuits, thus providing the high starting voltage that contributes to the success of the combination. When the power from the r-f generator is first applied to the helical resonator and the lamp has not yet started to discharge, the circuit has an impedance of 400 to 500 ohms and a very high “Q”, with a voltage maximum at the open-circuited end of the coil. The high voltage creates the potential that ionizes the gas in the lamp. When the discharge starts, the “Q” of the helical resonator drops, the voltage subsides, and the lamp maintains discharge through power provided in the turns of the coil. Should the discharge extinguish, by Gabriel’s design the voltage builds in the open-circuited end of the coil and the discharge spontaneously reignites. There is eliminated the intermediate switching or alternative power supply connections that were previously required to operate EDLs, as well as the necessity for adjustment in operation to maintain an impedance match.
The Westinghouse device works the same way. As stated in the ’884 patent, a device with these characteristics and advantages was not previously known, although helical resonators were well known. Further, the district court found that Westinghouse was unable to make a successful device until it adopted the details it learned by dismantling the Perkin-Elmer device. The majority has denigrated these facts. Although not controlling of the result, such facts are always to be considered in the application of equitable principles. Precedent binds us to apply equitable principles when required “[t]o temper unsparing logic and prevent an infringer from stealing the benefit of an invention”. Graver Tank & Manufacturing Co. v. Linde Air Products Co., 339 U.S. 605, 608, 70 S.Ct. 854, 856, 94 L.Ed. 1097 (1950).
The Claim Clauses
Each claim clause was explained by experts and made the subject of specific find*1539ings, variously contested on appeal. The majority does not discuss the claim limitations other than those of claim clauses (h) and (i). When the invention is considered in its entirety, as delineated in all of the claim clauses, clauses (h) and (i) are placed in perspective. To apply clauses (h) and (i) in anything other than the context of the entire claimed invention is improper claim analysis. Martin v. Barber, 755 F.2d 1564, 1568, 225 USPQ 233, 235 (Fed.Cir.1985); Hughes Aircraft Co. v. United States, 717 F.2d 1351, 1364, 219 USPQ 473, 482 (Fed. Cir.1983). The district court considered the entire invention, as must we.
Clause (a)
The district court found, without dispute on appeal, that the Westinghouse I and II devices are combinations of a helical resonator coupler and an EDL, Perkin-Elmer Corp., slip op. at 32, as stated in claim clause (a):
(a) A resonator coupler for coupling a source of r-f electric power into an electrodeless discharge lamp for starting and operating the lamp[.]
Clause (b)
It was not disputed that the Westinghouse I and II devices contain the element of claim clause (b):
(b) a grounded hollow cylinder of electrically conductive material open at one end, with a grounded base member at the other end[.]
Clause (c)
The district court found that the Westinghouse couplers “broadly correspond” to claim clause (c):
(c) a helically coiled wire conductor concentrically within the cylinder and spaced from the inner walls thereof[.]
Id. at 20.
Clause (d)
The district court found that there was a difference in the position of the Gabriel and the Westinghouse lamps, and thus concluded that clause (d) did not read on the accused devices:
(d) means for mounting a discharge lamp substantially concentrically within one end position of the coil[.]
Id. at 32. Perkin-Elmer assigns clear error to this finding, and I must agree. Westinghouse argued that its lamp is not wholly contained within the coil, but the claim does not so require and the specification does not so teach. The ’884 specification states at column 4, lines 50-54:
The position of the lamp relative to the coil is thus adjustable for finding the exact position at which the energy supplied by the coil ... is coupled into the lamp most efficiently when the lamp is operating.
It was uncontroverted that the Westinghouse lamp positions relative to the coil were adjustable, for this same reason: to match the impedance of the bulb and coupler to the energy source. It was also uncontroverted that in the Westinghouse devices at least a substantial portion of the lamp was within the helical coil, working in the same way as the Perkin-Elmer device. The district court clearly erred in its application of this clause.
Clause (e)
Claim clause (e) relates to the wave length of the resonator coil:
(e) the wire of the coil being one quarter wave long relative to the free-space wavelength of r-f power intended to be applied for operating a lamp mounted therein[.]
The district court found that Westinghouse I had a quarter-wavelength coil, id. at 18, within the plain scope of clause (e). In Westinghouse II some turns of the quarter-wavelength coil were removed and replaced with a capacitor. Although the district court gave cumulative weight to this change, as discussed infra, the court stated that the Westinghouse II modifications were “known prior art options for helical resonator circuits”. Id.
Clause (f)
Claim clause (f) provides:
*1540(f) the end of the coil at the end portion within which a lamp is adapted to be mounted being toward said base member and being grounded, the other end of the coil being open circuited[.]
Perkin-Elmer’s brief contained the following illustration from the record, showing the coil grounded at one end and open circuited at the other:
[[Image here]]
P-E
HEUCAL RESONATOR
The district court found that in the Westinghouse I and II couplers one end of the coil was connected to the r-f power source and not to ground. Id. at 20. As shown in the drawings, the Westinghouse device differs from the Perkin-Elmer device solely by the inversion of r-f power input and ground.
[[Image here]]
WESTINGHOUSE HELICAL RESONATOR
[Westinghouse I]
Westinghouse’s expert Hruda testified that the Westinghouse coil is grounded at the end of its quarter wavelength, as in claim clause (f), and the rest of the coil is a coupling loop to the r-f power. The district court clearly erred in giving no credence to this fundamental principle of electrical circuitry.
Clause (g)
It was not disputed that claim clause (g) reads on both Westinghouse I and II:
(g) electrical connecting means for connecting to the coil a source of r-f electrical power that is sufficient to maintain a discharge in a lamp mounted within the coil[.]
The use of radio frequency power to start and operate the lamp is fundamental to the claimed invention. Westinghouse did not challenge this point.
Clauses (k) and (i)
The majority bases its affirmance of non-infringement on claim clauses (h) and (i). The district court found that neither Westinghouse I nor II was connected to the r-f power supply in the same way as in these claim clauses:
(h) said connecting means being tapped into the coil at a point near, but spaced from, the grounded end thereof,
(i) said point being selected such that, when a lamp mounted in the coil is in operation by r-f power connected to the coil, the coupler is tuned to the frequency of said r-f power and the impedance of said lamp and coupling means at said tap *1541point substantially matches the impedance of said r-f power source[.]
Id. at 22-23. The court found that the Westinghouse devices lacked a tap point for the purposes of impedance matching and tuning. Id. at 23. The court also found that a coaxial cable used by Westinghouse between the coupler connection and the r-f source was an integral element of its system and was absent from the PerkinElmer teaching. Id.
Perkin-Elmer states, correctly, that these differences are not significant. The coaxial cable is described in the Perkin-Elmer specification as the “connecting means” of the claim. Perkin-Elmer further points out that clauses (h) and (i) describe a transformer connection between the r-f power source and the helical coil, shown in the prior art as interchangeable with the transformer connection used by Westinghouse. The following illustration is Figure 3 from Macalpine & Schildknecht, Helical Resonator Design Chart, 33 Electronics 140 (1960), and shows the loop transformer coupling of Westinghouse on the left side at (B) and the tap (auto)-transformer coupling of Perkin-Elmer on the right side at (B).
[[Image here]]
Perkin-Elmer’s argument that Macalpine shows this interchangeability of the transformer couplings to the helical resonators is a controlling factor in this analysis. This reference, used by both Westinghouse and Perkin-Elmer in the design of their respective devices, states:
Coupling into and out of the resonators can be effected by a tap, loop, probe or aperture in a manner analogous to that with straight quarter-wave coaxial lines.
Id. at 146. Although the district court, and the majority of this panel, erred in ignoring the Macalpine teaching, the position has not been dropped by Perkin-Elmer.
Westinghouse does not dispute that Macalpine shows the interchangeability of tap and loop transformer coupling to transfer power to the helical resonator. Rather, Westinghouse argues that its coupling does not rely on a tap point in order to achieve an impedance match. Westinghouse (and the majority) thus avoided, or obscured, the issue, by concentrating on a detail of the known difference between an autotransformer and a loop transformer. This difference is illustrated by Macalpine; and is of no moment to their use. It is the similarities between the transformers that are pertinent to this analysis. Both transformers perform the same function (electrical power transfer), in the same way (electromagnetic induction), with the same result (excitation of the resonator). Both transformers achieve an impedance match with the r-f power source. Indeed, Westinghouse’s expert Hruda testified that as to these couplings there “would be relatively little difference between Perkin-Elmer on our unit and our head on Perkin-Elmer’s unit”.
“It has often been said that a useful guide in adjudging the equivalency between a claimed and an accused device is whether persons reasonably skilled in the art would have known of the interchangeability of a part in the accused device with one in the claimed device.” Lockheed Aircraft Corp. v. United States, 553 F.2d 69, 82, 193 USPQ 449, 461 (Ct.Cl.1977). In this case the same transformer function is performed in a known interchangeable way with the same result.
It is of compelling weight that interchangeability of these specific transform*1542ers in this specific use is taught in the prior art. The record does not support the new interpretation of the principles of physics and law expounded by the majority.
Clause (j)
As viewed by the district court, claim clause (j), which states:
(j) whereby when said r-f power is applied to the coil, and before a discharge is ignited in the lamp, a voltage maximum occurs at the open circuited end of the coil and creates a potential extending through the lamp portion between said open circuited end and said base member for ionizing the gas in the lamp[,]
was avoided by Westinghouse because of the court’s finding that there is no voltage maximum in the accused devices. Perkin-Elmer Corp., slip op. at 22. This finding is clearly in error. Westinghouse had itself advised the PTO that claim clause (j) “merely recites the inherent function arising from the helical resonator structure.”
It is the voltage maximum that creates the potential that ionizes the gas in the lamp, as Westinghouse agreed. This is a known electrical characteristic of helical coils of the Macalpine type, used both in the Perkin-Elmer and the Westinghouse devices. When asked whether there was any difference between the Westinghouse and Perkin-Elmer devices with respect to the electrical phenomena by which they are started, Mr. Hruda responded: “No. Electromagnetic field starts them all.”
The District Court’s Analysis
The district court viewed the accused devices as a whole, and assessed the totality of their differences from the ’884 claims. The district court found that there were four structural differences between the Westinghouse devices and the ’884 claims, as discussed supra, which taken together the court found sufficient to avoid infringement. Three of these four differences appear to be based on a misunderstanding of the device, and the fourth on a misperception of the known equivalency of simple electrical structures.
The majority in focusing on only the fourth difference may have recognized the weaknesses in the decision relating to the other three, but by affirming the district court on less than that on which the court relied, the majority has departed even farther from our consistent requirement that the invention as a whole be considered. See, e.g., Martin, 755 F.2d at 1568, 225 USPQ at 235; Hughes Aircraft Co., 717 F.2d at 1364, 219 USPQ at 482; Carman Industries, Inc. v. Wahl, 724 F.2d 932, 942, 220 USPQ 481, 488-89 (Fed.Cir.1983); Tate Engineering, Inc. v. United States, 166 USPQ 329, 335 (Ct.Cl.1970).
The doctrine of equivalents does not authorize judicial revision of the basic invention; the interested public must know with reasonable precision the boundaries of the patent grant. See General Electric Co. v. Wabash Appliance Corp., 304 U.S. 364, 369, 58 S.Ct. 899, 901, 82 L.Ed. 1402 (1938); McClain v. Ortmayer, 141 U.S. 419, 423-24, 12 S.Ct. 76, 77, 35 L.Ed. 800 (1891). Nor is the purpose of the doctrine to cure errors curable by reissue, or to authorize judicial reexamination of claims. The purpose is to enable a court to remedy a “fraud on a patent”, in the words of Graver Tank, 339 U.S. at 608, 70 S.Ct. at 856, in order to achieve a just result.
Applying the standards established by this court, specifically the “function/way/result” test of Graver Tank, the Westinghouse transformer coupling meets all the requirements, as tested by precedent, of equivalency to that of claim clauses (h) and (i). This test has been affirmed and reaffirmed by this court. See, e.g., D.M.I., Inc. v. Deere & Co., 755 F.2d 1570, 1575, 225 USPQ 236, 239 (Fed.Cir.1985); Martin, 755 F.2d at 1568, 225 USPQ at 235; King Instrument Corp. v. Otari Corp., 767 F.2d 853, 862, 226 USPQ 402, 408 (Fed. Cir.1985), cert. denied, 106 S.Ct. 1197 (1986); Palumbo v. Don-Joy Co., 762 F.2d 969, 974-75, 226 USPQ 5, 8 (Fed.Cir.1985); Perkin-Elmer Corp. v. Computervision Corp., 732 F.2d 888, 900, 221 USPQ 669, 678 (Fed.Cir.), cert. denied, 469 U.S. 857, 105 S.Ct. 187, 83 L.Ed.2d 120 (1984).
*1543The Westinghouse devices meet all of the claim limitations. Nothing is “erased”, as charged by the majority. Carman Industries, cited by the majority for the proposition that the doctrine of equivalents applies when there is a “minor modification”, is squarely met in this case, where almost all the claim clauses are infringed literally, and one claim limitation is met by a known equivalent structure. Viewing the invention as a whole, infringement of claim 1 was clearly made out.
II
There is a presumption that a claim covers, and limits, the subject matter which “the applicant regards as his invention”, in the words of 35 U.S.C. § 112. When equitable principles are invoked in pursuit of fairness, to enlarge a claim’s reach beyond the applicant’s chosen words, there is a special burden on the courts. If courts are not to hinder the progress of technological advance, certainty and predictability are as important in the application of equitable as of legal principles.
The purpose of the doctrine of equivalents is to allow courts to do justice where justice requires. Yet it necessarily dilutes the rigorous certainty that patent claims are intended to provide. A fair balance between these conflicting purposes is not easily struck. We are carried into the tenuous area where equitable considerations favoring a patentee confront, head to head, the historic right and opportunity of scientists and engineers to design around the claims of a patented device.
Yet, unless is it decided by Congress, as a matter of public policy, that the decades of equity are to be revoked or circumscribed, it is not our role to do so. I see little value in the majority’s suggestion that inventors should rely on the reissue statute for relief, for this implies that relief is available from that source. The reissue (and reexamination) statutes can not serve this purpose. By reissue enlarged claims are not available after the patent is two years old, even if supported in the specification. 35 U.S.C. § 251. In no event is it possible to add to the specification. Neither reissue nor reexamination would have allowed Gabriel to add, even one day after his patent was issued, a reference in his specification to Fig. 3 or other disclosure of Macalpine, in order commensurately to broaden his claims.
Patent applicants need include in their patent .specifications no more of the prior art than is necessary to convey the invention to those of ordinary skill in the art, who are charged with knowledge of the prior art. See Lindemann Maschinenfabrik GMBH v. American Hoist and Derrick Co., 730 F.2d 1452, 1463, 221 USPQ 481, 489 (Fed.Cir.1984). Our predecessor court commented in In re Smythe, 480 F.2d 1376, 1384, 178 USPQ 279, 285 (CCPA 1973):
The alternative places upon patent applicants, the Patent Office, and the public the undue burden of listing, in the case of applicants, reading and examining, in the case of the Patent Office, and printing and storing, in the case of the public, descriptions of the very many structural or functional equivalents of disclosed elements or steps which are already stored in the minds of those skilled in the arts, ready for instant recall upon reading the descriptions of specific elements or steps.
That established law is useless, however, if courts refuse to recognize such dramatic “structural and functional equivalency” as is contained in the Macalpine article, illustrated supra.
Although it is my view that infringement litigation is not the best way to reach the technological judgments involved in equivalency determinations, the judicial system provides no ready alternative. One might fault Gabriel for failing to discuss or claim the known alternative prior art transformers, but the law was on his side, until today. The “semantic strait jacket” imposed on Perkin-Elmer by the majority, ignoring the prior art, is in derogation of our own precedent.
Failure of this court to reach consistent decisions based on a consistent application of precedent will be as destructive of the purposes of a patent system as was the forum-shopping and inconsistent judgments *1544of the past. I repeat my concern at the majority’s categorization in footnote 8 of much of this court’s jurisprudence as “dicta”. That message can not reassure those who hoped that this court would bring stability, not turmoil, to the patent law. If the majority wishes to redirect the judicial approach to equivalency, it does not write on a clean slate. That slate includes much more than this court’s decisions, and can no more be erased than can the clauses of Perkin-Elmer’s claims; although the law can surely be refined as new facts probe its boundaries. I suggest that this discussion by the majority is itself dicta — a luxury usually reserved to dissents.
The majority asserts that Gabriel did not describe his coupler as a "helical resonator”; Gabriel described his invention as a "resonator coupler” having a "helically coiled wire”, and the district court found “that by using a helical resonator [Gabriel] was able to achieve an essentially self-starting unit”. Id. at 11.