Appeal from a judgment of the United States District Court for the Eastern District of New York. The district court found that Westinghouse’s accused devices do not infringe claim 1 under the doctrine of equivalents. Unable to view that finding as clearly erroneous under Fed.R.Civ.P. 52(a), we affirm.
I.
Perkin-Elmer sued Westinghouse for infringing its U.S. Patent No. 3,873,884 (the ’884 patent) on an invention of Fred C. Gabriel (Gabriel), entitled “Electrodeless Discharge Lamp and Power Coupler Therefor”. A nonjury trial on the issues of patent infringement and validity was held on August 6-10, 1979.
Before the district court, Perkin-Elmer conceded the absence of literal infringement, but alleged infringement under the doctrine of equivalents. Westinghouse conceded that the ’884 patent, if narrowly construed, is valid. As noted by the district court, if the claim is to be “construed broadly as claimed by Perkin-Elmer, Westinghouse argue[d] that the patent is invalid for lack of specificity, lack of novelty, and obviousness.” Saying the patent is valid “if interpreted narrowly,” the district court upheld the validity of the ’884 patent and found that Westinghouse did not infringe under the doctrine of equivalents.
Perkin-Elmer appeals from the district court’s holding of non-infringement under the doctrine of equivalents. Westinghouse did not appeal from the district court’s validity holding.
II.
The sole issue is whether the district court’s finding, that the accused devices do not infringe claim 1 of the ’884 patent under the doctrine of equivalents, is clearly erroneous.1
III.
The claimed invention is a resonator coupler for an electrodeless discharge lamp (EDL).
EDL’s have a trace of a selected element incorporated in them to provide a bright spectral line source of the selected metal for use in chemical analysis, particularly atomic absorption spectroscopy (AAS). AAS is performed by comparing the intensity of the EDL-generated spectral line, after it has passed through a test sample which has been atomized in a flame or other atom cell, to a previously measured and stored intensity of the EDL-generated spectral line after it has passed through a sample of a known concentration which has been similarly atomized. Although EDL’s were known spectral line sources, they were of limited use due to their high cost, large size, unreliable operation, and undesirable electromagnetic problems.
Gabriel claimed his invention in very specific terms. The only independent claim in issue, claim 1, reads:
(a) 2 A resonator coupler for coupling a source of r-f electrical pcwer into an electrodeless discharge lamp for starting and operating the lamp comprising:
(b) a grounded hollow cylinder of electrically conductive material open at one end, with a grounded base member at the other end;
(c) a helically coiled wire conductor concentrically within the cylinder and spaced from the inner walls thereof;
(d) means for mounting a discharge lamp substantially concentrically within one end position of the coil;
*1530(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;
(f) the end of the coil at the end portion within which a lamp is adapted to be mounted being toward said base and being grounded, the other end of the coil being open circuited; and
(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,
(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 point substantially matches the impedance of said r-f power source,
(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. [Emphasis added.]
The parties disputed the meaning of several clauses in the claim. The district court did not expressly construe claim 1; it clung to the literal language of the claim and found four fundamental differences between the claimed invention and the accused products of Westinghouse. Because of those differences, the district court found that the accused devices do not perform substantially the same function in substantially the same way to obtain the same result.
It is unnecessary for this court in this case to ascertain the interpretation the district court gave to the disputed clauses in the claim, from a review of the differences it found in those clauses or otherwise.3 If that were done, and if we were then to conclude that the district court’s construction of the claim was legal error, we would remand to the district court for an application of the properly construed claim to Westinghouse’s devices.4 Because of the differences found by the district court between the accused devices and the structure and operation set forth in claim clauses (h) and (i), the interpretation of which is undisputed, and because those differences sufficiently support the district court’s determination of non-infringement under the doctrine of equivalents, a remand is not necessary and an affirmance is in order.5
*1531As often occurs, the record contains some evidence on which this court might, if sitting at trial, have found the facts differently. That is not, however, our role. When, as here, the district court has found the fact of non-equivalence and there is evidence to support that finding, we are not at liberty to reverse unless we can say that that finding was clearly erroneous. Anderson v. City of Bessemer City, 470 U.S. 564, 574, 105 S.Ct. 1504, 1512, 84 L.Ed.2d 518 (1985) (“Where there are two permissible views of the evidence, the fact finder’s choice between them cannot be clearly erroneous.”). In sum, a determination of non-equivalence is a finding of fact and we cannot say that that finding in this case was clearly erroneous.
Tap Coupling vs. Loop Coupling
The differences based on claim clauses (h) and (i) are for ease of reference denoted by the parties and the district court as tap coupling in the claimed invention and loop coupling in Westinghouse’s two devices.
The patent discloses and claims in clauses (h) and (i) an autotransformer-type tap coupling with a tap point variably located on the helical coil for connecting it to the r-f power source. To obtain a high voltage within the EDL, the location of the tap point is selected to tune the resonant frequency of the EDL to the frequency of the r-f power source. And to maximize power transferred from the r-f source to the EDL, the location of the tap point is set to match the internal impedance of the EDL with the impedance of the r-f power source. The patent specification, describing the single embodiment in the drawing and specifically set out in the claim, makes clear that the tap was located on the coil to perform a particular function:
The tap 26 into the coil 15 is located relative to the turns of the coil so that when the lamp 10 is operating the coupler 11 is tuned to the frequency supplied by the r-f generator and the impedance of the coupler and lamp circuit matches the impedance of the r-f generator at the tap 26.
The district court found that the tap point and its positioning for purposes of frequency tuning and impedance matching, as claimed, is one of the “cardinal structural-functional-operational interrelationships” of the claimed invention. Perkin-Elmer does not challenge that critical finding.
Perkin-Elmer concedes that Westinghouse’s devices employ a transformer-type loop coupling (as distinguished from an autotransformer-type tap coupling) in which the connecting point between the helical coil and the r-f power source is not fixed for purposes of frequency tuning or impedance matching. Perkin-Elmer also concedes that the claimed invention’s auto-transformer-type tap-coupling arrangement is different from Westinghouse’s transformer-type loop-coupling arrangement. In the claimed invention, a portion of the secondary coil in the autotransformer circuit serves as a primary coil in that circuit, whereas, in the accused devices, the primary and secondary coils in the transformer circuit are separate.
Perkin-Elmer’s candid concession of the aforementioned differences between the claimed invention and the accused devices is mandated by the record before us and there is no plausible basis for setting aside the findings on those differences as clearly erroneous. However, the critical question here is whether the differences in coupling establish that the claimed invention and the accused devices do not operate in substantially the same way.6 If that question is answered in the affirmative, the district court’s finding that the accused de*1532vices do not infringe claim 1 under the doctrine of equivalents cannot be found to have been clearly erroneous.
Perkin-Elmer’s Arguments
Perkin-Elmer argues that: (1) the claimed invention should be considered as comprised of only two elements, an EDL and a helical resonator, saying that the claimed invention is “more of a pioneering invention than merely an improvement” and is thus entitled to a broad range of equivalents, citing commercial success and industry impact; (2) the accused devices and the claimed invention are based on the same “concept”; and (3) tap and loop coupling were known by those with ordinary skill in the art to be interchangeable for power transfer in other devices.
(1) Two-element Combination
A pioneer invention is entitled to a broad range of equivalents. Sealed Air Corp. v. United States International Trade Comm’n, 645 F.2d 976, 984, 209 USPQ 469, 477 (CCPA 1981). However, the district court’s view that Perkin-Elmer’s claimed invention devoted to the provision of light for AAS is not a pioneer invention is supported by the presence in the record of prior art devices, including EDL’s, also devoted to provision of light for AAS. As specifically stated in the patent and undisputed on this record, the claimed invention constitutes an improved means for providing such light. Perkin-Elmer’s statement in its brief that the invention is “more in the nature of a pioneer patent [sic, invention] than a mere improvement” cannot substitute for evidence establishing that the invention was a pioneer. That an improvement enjoys commercial success and has some industry impact, as many do, cannot compel a finding that an improvement falls within the pioneer category.
Perkin-Elmer bases its “pioneer” argument on its repeated characterization of the claimed invention as a two-element combination of a known helical resonator with a known EDL.7 That characterization oversimplifies and disregards the specific language deliberately selected by Gabriel in claiming the subject matter he regarded as his invention. That Gabriel may have drawn knowledge from the helical resonator art is irrelevant, for the disclosure and claims of the ’884 patent are drawn to explicit structure and relationships having specific functions in EDL operation.
One must start with the claim, and,though a “non-pioneer” invention may be entitled to some range of equivalents, a court may not, under the guise of applying the doctrine of equivalents, erase a plethora of meaningful structural and functional limitations of the claim on which the public is entitled to rely in avoiding infringement. Nor, as Perkin-Elmer here requests, should a court convert a multi-limitation claim to one of two limitations to support a finding of equivalency. If Gabriel regarded his invention to have been as Perkin-Elmer now describes it, he should have at least referred to a helical resonator in his disclosure and either so claimed it or sought broadened claims by reissue. Though the doctrine of equivalents is designed to do equity, and to relieve an inventor from a semantic strait jacket when equity requires, it is not designed to permit wholesale redrafting of a claim to cover nonequivalent devices, i.e., to permit a claim expansion that would encompass more than an insubstantial change. Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 610, 70 S.Ct. 854, 857, 94 L.Ed. 1097, 85 USPQ 328, 331 (1950); Carmen Industries v. Wahl, 724 F.2d 932, 942, 220 USPQ 481, 488 (Fed.Cir.1983) (doctrine applies when there is a “minor modification”).
In Hughes Aircraft Co. v. United States, 717 F.2d 1351, 1364, 219 USPQ 473, 482 (Fed.Cir.1983), this court noted that it was legal error not to “apply the doctrine *1533of equivalents to the claimed invention as a whole.” That statement dealt with an infringement inquiry implicating an entire claim, as distinguished from a section 112 ¶ 6 inquiry implicating only a “means plus function” limitation of a claim. That statement also was a recognition that, in applying the doctrine of equivalents, each limitation must be viewed in the context of the entire claim. The statement should not be interpreted as sanctioning the treatment of claim limitations as insignificant or immaterial in determining infringement.8 “It is ... well settled that each element of a claim is material and essential, and that in order for a court to find infringement, the plaintiff must show the presence of every element or its substantial equivalent in the accused device.” Lemelson v. United States, 752 F.2d 1538, 1551, 224 USPQ 524, 533 (Fed.Cir.1985).9 To be a “substantial equivalent,” the element substituted in the accused device for the element set forth in the claim must not be such as would substantially change the way in which the function of the claimed invention is performed.
Perkin-Elmer’s repeated insistence on interpreting claim 1 as though it read “In combination, an EDL and a helical resonator” is accompanied by no explanation of why the invention was never so claimed. The major portion of its brief is devoted to a review of the prior art in an effort to indicate that such a claim might have been allowed. Because it was never submitted, we are left only with conjecture on what the examiner’s reaction might have been and what prior art might have been cited against such a claim.
Moreover, the accused devices were on the market for a year when Perkin-Elmer filed for “reissue” under 37 C.F.R. § 1.175(a)(4)(1977). Westinghouse participated actively in that proceeding, during which Perkin-Elmer succeeded in establishing that claim 1, os presently written, distinguished over numerous additional prior art references. Thus, with knowledge of the accused devices, Perkin-Elmer at no time elected to submit its two-limitation claim to PTO examination.10
Throughout its briefs, Perkin-Elmer tells us often that Westinghouse as protester said that Gabriel’s invention was just an old EDL and an old helical resonator. Doubtless, as protestors will, Westing*1534house sought to broaden the claim to the point of invalidity. Perkin-Elmer’s effort to cite Westinghouse’s protest-inspired characterization as authority or evidence is twice flawed: (1) it lacks any statement of what Perkin-Elmer said to the PTO in response; and (2) it fails to realize that the examiner did not accept Westinghouse’s characterization but simply found the detailed claims as written not unpatentable over the art.
Here, as above indicated, the district court correctly rejected Perkin-Elmer’s two-element characterization of the claimed invention. The district court properly viewed the entire claim, including clauses (h) and (i), and found that the accused devices did not operate in substantially the same way as did the claimed invention. Perkin-Elmer’s attempt at minimizing the importance of clauses (h) and (i) cannot establish clear error in the district court’s findings 53 and 54:
53. Apart from the absence of the tap point, the impedance of the Westinghouse power coupler at the point of r-f connection does not substantially match the impedance of the r-f source. In Westinghouse couplers, there is an impedance mismatch and the coaxial cable between the coupler connection and the r-f source is an integral element used for impedance matching the remote r-f source to the system. Such cozxial cable must be of a specific length to cancel out the inductance provided by the transformer coupling loop which presents an additional inductive reactance not present in the circuit described and claimed in the Gabriel patent.
54. The Westinghouse power couplers are not tuned by anything relating to the internal r-f power coupling means. Dr. Nahemow pointed out that the tuning of the Westinghouse power coupler is done by adjusting the length of the coaxial cable between the connecting means and the r-f power source and by positioning the iris capacitor. [Citations omitted.]
(2) Concept
In attempting to denigrate the significance of the differences between tap-coupling and loop-coupling in this case, PerkinElmer says both methods use transformer “concepts” to transfer power to an EDL. Again, as occurs throughout its brief, Per-kin-Elmer improperly generalizes the claimed invention. The claim is not drawn, as no claim may be, to a “concept”, and is not directed merely to power transfer. The claimed tap coupling accomplishes much more than merely transferring power. As the district court found, tap-coupling enables the claimed invention to implement a specific manner or way of impedance matching and frequency tuning. That manner of operation is based on a positioning of the tap point location along the length of the helical coil, and is, as PerkinElmer concedes and the district court stated supra, one of the “cardinal structural-functional-operational interrelationships” of the claimed invention.
In contrast, there is no such tap point in Westinghouse’s devices, for impedance matching, for frequency tuning, or for any purpose.
There is on the contrary, as the district court found, a frequency mismatch in the Westinghouse devices which is remedied by varying the length of an external cable (not an element in the claimed invention) connecting the EDLs to their r-f power source and by the positioning of an “iris capacitor”; that is, the Westinghouse devices are not tuned by anything relating to the internal r-f power coupling means and are tuned in a substantially different way from that required by the claim. Perkin-Elmer does not challenge the district court’s findings of those structural and functional differences.
The impedance of Westinghouse’s devices at the point of r-f connection does not substantially match the impedance of the r-f source. Rather, there is, in addition to the above-described frequency mismatch, an impedance mismatch. That mismatch is in part caused by an additional inductive reactance (impedance) generated by Westinghouse’s use of loop coupling. Those facts are undisputed.
*1535The district court found that, to compensate for that impedance mismatch, the external cable connecting Westinghouse’s EDLs to their r-f power source must be of a specific length, and is an “integral element used for impedance matching.” Per-kin-Elmer’s response to that finding is to argue that the invention and the accused devices all match impedance. That argument misses the mark, for it disregards the way impedance is matched in its claimed invention and the different way impedance is matched in the accused devices.
Perkin-Elmer says an impedance match could be achieved in the accused devices by varying the number of turns in the primary coil of their transformer circuit, that primary coil being the portion of helical coil between the fixed tap-point at one end and the ground connection. Assuming, arguendo, that that were true, we fail to detect the relevance of a modified version of the accused devices, even if Perkin-Elmer had pointed, as it did not, to evidence that the modified version would operate substantially the same as the claimed invention. That a hypothetical modification of the accused devices might make them more nearly equivalents of the claimed invention argues against a finding of infringement by the actual accused devices under the doctrine of equivalents.
(3) Interchangeability
That persons skilled in the art would have known of the interchangeability of claimed with unclaimed elements is a factor in considering equivalence, Graver Tank & Mfg. Co., 339 U.S. at 609, 70 S.Ct. at 856, 85 USPQ at 331; see Thomas & Betts Cory., 720 F.2d at 1579, 220 USPQ at 6, yet the accused devices must still perform substantially the same function in substantially the same way to obtain the same result. The present record is replete with evidence directly establishing the substantially different ways in which the claimed resonator coupling and the accused devices operate. The district court’s findings on that evidence are for the most part unchallenged. To the extent Perkin-Elmer does challenge the district court’s findings on the operational differences of the coupling arrangements, those challenges are vague, ambiguous, and incomplete. For example, Perkin-Elmer’s evidence that tap-coupling and loop-coupling were known to be interchangeably useful in effecting power transfer in entirely different and unrelated environments cannot serve as a basis for enlarging the subject matter explicitly set forth in the claim. Moreover, that evidence is nonprobative of equivalence when one notes that the coupling structure of the claimed invention does more than merely transfer power. As above indicated, the claimed tap coupling also assumes a critical role in matching impedance and tuning frequency, functions achieved in an entirely different way in the accused devices. Perkin-Elmer points to no evidence that tap and loop couplings were known as interchangeable means for performing those functions.
That the functions of impedance matching and frequency tuning are not performed in substantially the same way establishes in this case that the accused devices do not wholly operate in substantially the same way as the claimed invention in providing light for AAS. Though the claimed invention and the accused devices provide light for AAS, the specific coupling structure and method specifically claimed permeates the entire operation of the claimed invention and cannot be deemed insignificant or insubstantial.
Conclusion
The district court’s finding that the accused devices and the claimed invention do not operate in substantially the same way cannot be said to have been clearly erroneous. Accordingly, the judgment appealed from is affirmed.
AFFIRMED.
. That Westinghouse purchased and studied a patented lamp of Perkin-Elmer and then designed its own is of no moment if its design does not constitute infringement.
. Lettering of claim 1 was added by district court, which we continue for convenience.
. Nor is it necessary to discuss the portion of the district court’s opinion devoted to the prosecution history. The detailed claims were allowed in the first Office Action. Whether the district court's discussion of the prosecution history may have been too succinct is irrelevant when, as here, its determination of non-equivalence between the accused devices and the claimed invention is not clearly erroneous. Perkin-Elmer’s argument respecting prosecution history estoppel disregards the effect of attorney argument and the basic requirement that claims be drawn to that which the inventor considers his invention, 35 U.S.C. § 112, and to avoid the prior art. In response to a rejection, for example, Gabriel’s attorney said:
All applicant’s claims define the structure by which his desired mode of operation is achieved and which distinguishes over the pri- or art exemplified by the foregoing Booth et at patent. Specifically, for example, applicant’s claims specify that his coil has one end “open circuited” and that means connecting the coil to the rf source is tapped into the coil at "a point near, but spaced from, the grounded end thereof’ (claims 1 to 10 or "intermediate the ends thereof' (claims 11 and 12)). [Emphasis added.]
. Validity of the properly construed claim would be a potential issue on such a remand.
. In SRI International v. Matsushita Electric Corp., 775 F.2d 1107, 227 USPQ 577 (Fed.Cir.1985) (in banc), this court reversed a grant of summary judgment because of the presence of genuine issues of material fact relating to the reverse doctrine of equivalents. The present case arises after a full trial on the application of the doctrine of equivalents. Whether differences were correctly found by the district court between the accused devices and other claim clauses is irrelevant where, as here, the differences in clauses (h) and (i) are such as to cause the accused devices to operate as wholes in a way not substantially the same as that in which the claimed devices operate as wholes.
. Perkin-Elmer’s repeated assertions that the claimed and accused devices perform substantially the same function and achieve substantially the same end result are not helpful. That circumstance is commonplace when the devices are sold in competition. That a claimed invention and an accused device may perform substantially the same function and may achieve the same result will not make the latter an infringement under the doctrine of equivalents where it performs the function and achieves the result in a substantially different way. Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 608, 70 S.Ct. 854, 856, 94 L.Ed. 1097, 85 USPQ 328, 330 (1950); see, e.g., Sealed Air Corp. v. U.S. International Trade Comm’n, 645 F.2d 976, 984, 209 USPQ 469, 476 (CCPA 1981).
. The words “helical resonator" appear nowhere in the patent and Gabriel’s invention is described nowhere in the prosecution history as incorporating a helical resonator. Gabriel specifically described and specifically claimed in detail a single embodiment of a particular resonator coupler, with each of its structural elements, relationships, and functions expressed in the claim. Nothing in the patent provides support for expanding the scope of that claim.
. In determining priority of invention, consideration of the "gist” or "essence” of the invention may be appropriate. See, e.g., Stansbury v. Bond, 482 F.2d 968, 974, 179 USPQ 88, 92 (CCPA 1973); McCutchen v. Oliver, 367 F.2d 609, 611, 151 USPQ 387, 389-92 (CCPA 1966); Hall v. Taylor, 332 F.2d 844, 848, 141 USPQ 821, 824 (CCPA 1964). We are aware of dicta that state consideration of the "essence", "gist", or "heart” of the invention may be helpful in determining infringement under the doctrine of equivalents. Loctite Corp. v. Ultraseal Ltd., 781 F.2d 861, 228 USPQ 90 (Fed.Cir.1985); Atlas Powder Co. v. E.I. Du Pont De Nemours & Co., 750 F.2d 1569, 1582, 224 USPQ 409, 418 (Fed.Cir.1984) (both citing Medtronic, Inc. v. Cardiac Pacemakers, Inc., 721 F.2d 1563, 1567, 220 USPQ 97, 101 (Fed.Cir. 1983). That dicta may not be read as implying that specific claim limitations can be ignored as insignificant or immaterial in determining infringement. It must be read as shorthand for the considerations set forth in Graver Tank, i.e., that the infringer should not appropriate the invention by making substitutions for those limitations, when the substitutions do not substantially change the function performed, or the way it is performed, by the invention.
. References to “elements” can be misleading. "Elements” often is used to refer to structural parts of the accused device or of a device embodying the invention. "Elements” is also used in the phrase "[a]n element of a claim” in 35 U.S.C. § 112 ¶ 6. An element of an embodiment of the invention may be set forth in the claim (e.g., "said connecting means” in clause (h) of the present claim). It is the limitation of a claim that counts in determining both validity and infringement, and a limitation may include descriptive terms (e.g., "tapped into the coil at a point near, but spaced from, the grounded end thereof’ in clause (h)). Because claims are composed of a number of limitations, the limitations have on occasion been referred to as "claim elements” or “elements of the claim,” but clarity is advanced when sufficient wording is employed to indicate when "elements” is intended to mean a component of an accused device or of an embodiment of an invention and when it is intended to mean a feature set forth in or as a limitation in a claim.
. Perkin-Elmer filed its (a)(4) “reissue” proceeding two years and five months after the '884 patent issued. To submit its broadened two-limitation claim it would have had to have filed for reissue within two years of the patent’s issue date. 35 U.S.C. § 251.