Telegenix, Inc. (“Telegenix”) appeals from the final judgment of the United States District Court for the Northern District of Texas in favor of Texas Digital Systems, Inc. (“TDS”). Because the district court erroneously construed certain disputed claim limitations, but correctly construed other claim limitations, we affirm-in-part, reverse-in-part, and remand.

BACKGROUND

TDS is the current owner of the four patents at issue, U.S. Patent Nos. 4,845,-481. (“'481 patent”), 4,965,561 (“'561 patent”), 4,734,619 (“'619 patent”), and 4,804,-890 (“'890 patent”), each issued to Karel Havel. TDS obtained these patents from Havel in 1997.

The Havel patents are directed to methods and devices for controlling the color of pixels in a light emitting diode (“LED”) display. Each pixel includes at least two elements corresponding to different pri[*1198] mary colors, e.g., one red element and one green element. Light signals from the two elements may be blended to produce a composite light signal of variable color. Figure 1 of the '481 patent, reproduced below, shows seven pixels arranged m a familiar seven-segment display pattern, each pixel having a red element (i.e., 2a-2g) and a green element (i.e., 3a-3g).

[[Image here]]

Claims 1 and 2 of the '481 patent are representative of the asserted claims of the '481 and '561 patents:

1. A method for controlling a color of a variable color display device which comprises a plurality of display areas arranged in a pattern for selectively exhibiting a plurality of display units, each said display area including a plurality of light sources for emitting upon activation light signals of respectively different primary colors and means for combining said light signals to obtain a composite light signal of a composite color, by exhibiting a selected display unit by repeatedly substantially simultaneously activating the light sources in selected display areas for brief time intervals to cause the light sources to emit light signals of said primary colors, and by selectively controlling the durations of the time intervals of activation of the light sources in the selected display areas to control the portions of the primary color light signals emitted therefrom, to thereby control the color of the exhibited display unit, (emphases added)

2. A variable color display device comprising:

a plurality of variable color display areas arranged in a pattern for selectively exhibiting a plurality of display units, each said display area including a plurality of light sources or emitting upon activation light signals of respectively different primary colors and means for combining said light signals to obtain a composite light signal of a composite color;

[*1199] means for exhibiting a selected display unit by repeatedly substantially simultaneously activating the light sources in selected display areas by pulses of a substantially constant amplitude for causing the light sources to emit light signals of said primary colors; and

color control means for selectively controlling the durations of the pulses applied to the light sources in the selected display areas to control the portions of the primary color light signals emitted therefrom, to thereby control the color of the exhibited display unit, (emphases added)

The '619 patent is directed to display devices including a variable color background area 82 substantially surrounding the display area segments 31, as illustrated in Fig. la, reproduced below.

[[Image here]]

Claim 1 of the '619 patent is representative of the claims and is reproduced below:

1. A variable color display device comprising:

a plurality of variable color display areas arranged in a pattern, each said display area including a plurality of display light sources for emitting upon activation light signals of different colors and means for combining said light signals to obtain a composite light signal of a composite color;

a variable color background area substantially surrounding said display areas and including a plurality of background regions adjacent to said display areas, each said background region including a plurality of light sources for emitting upon activation light signals of different colors and means for combining said light signals to obtain a composite light signal of a composite color;

a plurality of opaque walls for optically separating said background regions from adjacent display areas; and

means for selectively activating said display light sources, to illuminate certain of said display ar'eas in a first color, and said background light sources, to illuminate said background regions in a second color different from said first color, (emphases added)

The '890 patent is directed to a variable color LED display and display circuit as illustrated in Figure 3, reproduced below:

[*1200] [[Image here]]

Representative claim 4 of the '890 patent is reproduced below:

4. A display device comprising:

a plurality of variable color display areas arranged in a pattern for selectively exhibiting a plurality of display units, each said display area including a plurality of light sources for emitting upon activation light signals of different colors and means for combining said light signals to obtain a composite light signal of a composite color;

first means for carrying selective display color control signals;

converter means for converting said display color control signals to obtain complementary color control signals;

second means for carrying said complementary color control signals; and[*1201] control means for selectively coupling said light sources in said display areas to said first means, for causing selective ones of said display areas to illuminate in a selected color defined by said display color control signals, and to said second means, for causing the remaining display areas to illuminate in a substantially complementary color defined by said complementary color control signals. (emphases added)

After TDS obtained the Havel patents in 1997, TDS filed suit, alleging that Telege-nix’s Colorgraphix devices infringed each of them. Following a jury verdict in favor of TDS, the district court entered judgment that Telegenix had literally infringed claims 1-4 and 7 of the '481 patent, claims 1-4 of the '561 patent, claim 1 of the '619 patent, and claim 4 of the '890 patent. The district court also found each of the asserted claims not invalid and concluded that Telegenix had willfully infringed “one or more” of the four asserted patents.

The district court awarded TDS a reasonable royalty of 20% as applied to $30 million in infringing sales (i.e., $6 million), enhanced damages of $6 million, pre-judgment interest of $3,007,999, post-judgment interest at 6.5%, and costs. The district court also permanently enjoined Telegenix from making, using, selling, or offering to sell its Colorgraphix color display devices, versions of its software used with the Co-lorgraphix color display devices, or other devices that otherwise infringe.

Telegenix appeals. We have jurisdiction pursuant to 28 U.S.C. § 1295(a)(1).

ANALYSIS

Standard of Review

Claim construction is a question of law that this court reviews de novo. Cybor Corp. v. FAS Techs., Inc., 138 F.3d 1448, 1456, 46 USPQ2d 1169, 1174 (Fed. Cir.1998) (en banc). The standard of review for jury instructions is prejudicial legal error. See Jamesbury Corp. v. Litton Indus. Prods., 756 F.2d 1556, 1558, 225 USPQ 253, 255 (Fed.Cir.1985), overruled on other grounds by A.C. Aukerman Co. v. R.L. Chaides Constr. Co., 960 F.2d 1020, 22 USPQ2d 1321 (Fed.Cir.1992) (en banc). To prevail, the party challenging the jury instruction “must demonstrate both that the jury instructions actually given were fatally flawed and that the requested instruction was proper and could have corrected the flaw.” Biodex Corp. v. Loredan Biomedical, Inc., 946 F.2d 850, 862, 20 USPQ2d 1252, 1261 (Fed.Cir.1991). “An erroneous instruction regarding claim interpretation that affects the jury’s decision on infringement is grounds for a new trial.” Ecolab Inc. v. Paraclipse, Inc., 285 F.3d 1362, 1373, 62 USPQ2d 1349, 1356 (Fed.Cir.2002).

Telegenix, which timely objected to the jury instructions at trial, argues that the district court erroneously interpreted the claims of the asserted patents in its Mark-man order and instructed the jury according to the erroneous claim constructions. Telegenix further argues that the district court abused its discretion in excluding certain evidence offered by Telegenix and in admitting other evidence presented by Texas Digital, and erroneously relied on the rule of Wine Railway Appliance Co. v. Enterprise Railway Equipment Co., 297 U.S. 387, 56 S.Ct. 528, 80 L.Ed. 736 (1936). On these grounds, Telegenix seeks a new trial. We address each of the allegations of error in turn.

I. The Contours of Claim Construction

“In construing claims, the analytical focus must begin and remain centered on the language of the claims themselves, for it is that language that the patentee chose to use to ‘particularly point[ ] out[*1202] and distinctly claim[ ] the subject matter which the patentee regards as his invention.’ 35 U.S.C. § 112, ¶2.” Interactive Gift Express, Inc. v. Compuserve, Inc., 256 F.3d 1323, 1331, 59 USPQ2d 1401, 1406 (Fed.Cir.2001). The terms used in the claims bear a “heavy presumption” that they mean what they say and have the ordinary meaning that would be attributed to those words by persons skilled in the relevant art. See CCS Fitness, Inc. v. Brunswick Corp., 288 F.3d 1359, 1366, 62 USPQ2d 1658, 1662 (Fed.Cir.2002); K-2 Corp. v. Salomon S.A., 191 F.3d 1356, 1362-63, 52 USPQ2d 1001, 1004 (Fed.Cir.1999); Johnson Worldwide Assocs., Inc. v. Zebco Corp., 175 F.3d 985, 989, 50 USPQ2d 1607, 1610 (Fed.Cir.1999); Specialty Composites v. Cabot Corp., 845 F.2d 981, 986, 6 USPQ2d 1601, 1604 (Fed.Cir.1988). Moreover, unless compelled otherwise, a court will give a claim term the full range of its ordinary meaning as understood by persons skilled in the relevant art. See Rexnord Corp. v. Laitram Corp., 274 F.3d 1336, 1342, 60 USPQ2d 1851, 1854 (Fed.Cir.2001); Johnson Worldwide Assocs., 175 F.3d at 989, 50 USPQ2d at 1610; Specialty Composites, 845 F.2d at 986, 6 USPQ2d at 1604.

It has been long recognized in our precedent and in the precedent of our predecessor court, the Court of Customs and Patent Appeals, that dictionaries, encyclopedias and treatises are particularly useful resources to assist the court in determining the ordinary and customary meanings of claim terms. See Teleflex, Inc. v. Ficosa N. Am. Corp., 299 F.3d 1313, 1325, 63 USPQ2d 1374, 1380 (Fed.Cir.2002) (“The ordinary meaning of a claim term may be determined by reviewing a variety of sources, including ... dictionaries and treatises .... ” (internal citations omitted)); CCS Fitness, 288 F.3d at 1366, 62 USPQ2d at 1662 (“[0]ur precedents show that dictionary definitions may establish a claim term’s ordinary meaning.”); Optical Disc Corp. v. Del Mar Avionics, 208 F.3d 1324, 1334-35, 54 USPQ2d 1289, 1295 (Fed.Cir.2000) (“For such ordinary meaning, we turn to the dictionary definition of the term.”); Quantum Corp. v. Rodime, PLC, 65 F.3d 1577, 1581, 36 USPQ2d 1162, 1166 (Fed.Cir.1995) (“[W]e see no error in the district court’s use of dictionary definitions to ascertain the ordinary meaning of the relevant claim limitation.”); In re Ripper, 36 C.C.P.A. 743, 171 F.2d 297, 299, 80 USPQ 96, 98 (1948) (“[I]t is clear that in ascertaining the meaning of [the claim term] as it appears herein, reference properly may be made to the ordinary dictionaries.”).

Dictionaries are always available to the court to aid in the task of determining meanings that would have been attributed by those of skill in the relevant art to any disputed terms used by the inventor in the claims. See Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1584 n. 6, 39 USPQ2d 1573, 1578 n. 6 (Fed.Cir.1996) (“[TJechnical treatises and dictionaries ... are worthy of special note. Judges are free to consult such resources at any time ... and may also rely on dictionary definitions when construing claim terms....”); Cybor Corp., 138 F.3d at 1459, 46 USPQ2d at 1177 (citing Vitronics for the proposition that a court is free to consult dictionaries, encyclopedias, and treatises at any time to help determine the meaning of claim terms); Vanguard Prods. Corp. v. Parker Hannifin Corp., 234 F.3d 1370, 1372, 57 USPQ2d 1087, 1089 (Fed.Cir.2000) (“A dictionary is not prohibited extrinsic evidence, and is an available resource of claim construction.”).

When a patent is granted, prosecution is concluded, the intrinsic record is fixed, and the public is placed on notice of its allowed claims. Dictionaries, ency[*1203] clopedias and treatises, publicly available at the time the patent is issued, are objective resources that serve as rehable sources of information on the established meanings that would have been attributed to the terms of the claims by those of skill in the art. Such references are unbiased reflections of common understanding not influenced by expert testimony or events subsequent to the fixing of the intrinsic record by the grant of the patent, not colored by the motives of the parties, and not inspired by litigation. Indeed, these materials may be the most meaningful sources of information to aid judges in better understanding both the technology and the terminology used by those skilled in the art to describe the technology.

These materials serve as important resources to assist courts in many ways. For example, they are often used to aid in the interpretation of statutes and regulations and in the interpretation of terms used in contracts. See, e.g., Rocknel Fastener, Inc. v. United States, 267 F.3d 1354, 1356-57 (Fed.Cir.2001) (advising that the interpretation of tariff terms, a matter of statutory construction, may be aided by reviewing “dictionaries, scientific authorities, and other reliable information sources” (citations omitted)); Am. Express Co. v. United States, 262 F.3d 1376, 1381 n. 5 (Fed.Cir.2001) (in interpreting Internal Revenue Service regulations and procedures, “[i]t is appropriate to consult dictionaries to discern the ordinary meaning of a term not explicitly defined by statute or regulation”); Bowers v. Baystate Techs., 302 F.3d 1334, 1345-46 (Fed.Cir.2002) (construing contract terms using non-technical and technical dictionaries); Buchanan v. Dep’t of Energy, 247 F.3d 1333, 1339 (Fed.Cir.2001) (relying on a dictionary definition in construing a settlement agreement). These materials deserve no less fealty in the context of claim construction.

As resources and references to inform and aid courts and judges in the understanding of technology and terminology, it is entirely proper for both trial and appellate judges to consult these materials at any stage of a litigation, regardless of whether they have been offered by a party in evidence or not. Thus, categorizing them as “extrinsic evidence” or even a “special form of extrinsic evidence” is misplaced and does not inform the analysis.

Because words often have multiple dictionary definitions, some having no relation to the claimed invention, the intrinsic record must always be consulted to identify which of the different possible dictionary meanings of the claim terms in issue is most consistent with the use of the words by the inventor. See Dow Chem. Co. v. Sumitomo Chem. Co., 257 F.3d 1364, 1372-73, 59 USPQ2d 1609, 1614 (Fed.Cir.2001); Multiform Desiccants, Inc. v. Medzam, Ltd., 133 F.3d 1473, 1478, 45 USPQ2d 1429, 1433 (Fed.Cir.1998). If more than one dictionary definition is consistent with the use of the words in the intrinsic record, the claim terms may be construed to encompass all such consistent meanings. Rexnord, 274 F.3d at 1343, 60 USPQ2d at 1858 (holding that the claim term “portion” may be interpreted in accordance with the dictionary definitions to encompass both “separate” and “integral” parts of an object). The objective and contemporaneous record provided by the intrinsic evidence is the most reliable guide to help the court determine which of the possible meanings of the terms in question was intended by the inventor to particularly point out and distinctly claim the invention. See Renishaw PLC v. Marposs Societa’ per Azioni, 158 F.3d 1243, 1250, 48 USPQ2d 1117, 1122 (Fed.Cir.1998) (“The construction that stays true to the claim language and most naturally[*1204] aligns with the patent’s description of the invention will be, in the end, the correct construction.”).

Moreover, the intrinsic record also must be examined in every case to determine whether the presumption of ordinary and customary meaning is rebutted. See id. Indeed, the intrinsic record may show that the specification uses the words in a manner clearly inconsistent with the ordinary meaning reflected, for example, in a dictionary definition. In such a case, the inconsistent dictionary definition must be rejected. See id. (“[A] common meaning, such as one expressed in a relevant dictionary, that flies in the face of the patent disclosure is undeserving of fealty.”); Lie bscher v. Boothroyd, 46 C.C.P.A. 701, 258 F.2d 948, 951, 119 USPQ 133, 135 (1958) (“Indiscriminate reliance on definitions found in dictionaries can often produce absurd results.”). In short, the presumption in favor of a dictionary definition will be overcome where the patentee, acting as his or her own lexicographer, has clearly set forth an explicit definition of the term different from its ordinary meaning. See In re Paulsen, 30 F.3d 1475, 1480, 31 USPQ2d 1671, 1674 (Fed.Cir.1994); Intellicall, Inc. v. Phonometrics, Inc., 952 F.2d 1384, 1387-88, 21 USPQ2d 1383, 1386 (Fed.Cir.1992). Further, the presumption also will be rebutted if the inventor has disavowed or disclaimed scope of coverage, by using words or expressions of manifest exclusion or restriction, representing a clear disavowal of claim scope. See Teleflex, 299 F.3d at 1324, 63 USPQ2d at 1380.

Consulting the written description and prosecution history as a threshold step in the claim construction process, before any effort is made to discern the ordinary and customary meanings attributed to the words themselves, invites a violation of our precedent counseling against importing limitations into the claims. See, e.g., Generation II Orthotics Inc. v. Medical Technology Inc., 263 F.3d 1356, 1367, 59 USPQ2d 1919, 1928 (Fed.Cir.2001) (“The district court should have construed the claim limitation ‘controlled’ according to its ordinary and accustomed meaning [citing medical dictionary], rather than importing a characteristic of a disclosed or preferred embodiment into that term.”); Loctite Corp. v. Ultraseal Ltd., 781 F.2d 861, 867, 228 USPQ 90, 93 (Fed.Cir.1985) (“Generally, particular limitations or embodiments appearing in the specification will not be read into the claims.”), overruled on other grounds by Nobelpharma AB v. Implant Innovations, Inc., 141 F.3d 1059, 46 USPQ2d 1097 (Fed.Cir.1998). For example, if an invention is disclosed in the written description in only one exemplary form or in only one embodiment, the risk of starting with the intrinsic record is that the single form or embodiment so disclosed will be read to require that the claim terms be limited to that single form or embodiment. See Teleflex, 299 F.3d at 1328, 63 USPQ2d at 1383 (“To the extent that the district court construed the term ‘clip’ to be limited to the embodiment described in the specification, rather than relying on the language of the claims, we conclude that the district court construed the claim term ‘clip (28)’ too narrowly.”); Comark Communications, Inc. v. Harris Corp., 156 F.3d 1182, 1186, 48 USPQ2d 1001, 1005 (Fed.Cir.1998) (cautioning against the limitation of the claimed invention to preferred or specific embodiments or examples); Transmatic, Inc. v. Gulton Indus., Inc., 53 F.3d 1270, 1277, 35 USPQ2d 1035, 1040-41 (Fed.Cir.1995) (“[A] patent claim is not necessarily limited to a preferred embodiment disclosed in the patent.”); SRI Int’l, Inc. v. Matsushita Elec. Corp., 775 F.2d 1107, 1121 n. 14, 227 USPQ 577, 585 n. 14 (Fed.Cir.1985) (en banc) (“That a specification describes[*1205] only one embodiment does not require that each claim be limited to that one embodiment”). Indeed, one can easily be misled to believe that this is precisely what our precedent requires when it informs that disputed claim terms should be construed in light of the intrinsic record. See, e.g., Markman v. Westview Instruments, Inc., 52 F.3d 967, 980, 34 USPQ2d 1321, 1329-30 (Fed.Cir.1995) (stating the claims must be construed in light of the specification and the patent’s prosecution history, if in evidence). But if the meaning of the words themselves would not have been understood to persons of skill in the art to be limited only to the examples or embodiments described in the specification, reading the words in such a confined way would mandate the wrong result and would violate our proscription of not reading limitations from the specification into the claims. See, e.g., Teleflex, 299 F.3d at 1328, 63 USPQ2d at 1383; Generation II Orthotics, 263 F.3d at 1367, 59 USPQ2d at 1928; Comark, 156 F.3d at 1186, 48 USPQ2d at 1005; Transmatic, 53 F.3d at 1277, 35 USPQ2d at 1040-41; SRI Int'l 775 F.2d at 1121 n. 14, 227 USPQ at 585 n. 14.

By examining relevant dictionaries, encyclopedias and treatises to ascertain possible meanings that would have been attributed to the words of the claims by those skilled in the art, and by further utilizing the intrinsic record to select from those possible meanings the one or ones most consistent with the use of the words by the inventor, the full breadth of the limitations intended by the inventor will be more accurately determined and the improper importation of unintended limitations from the written description into the claims will be more easily avoided.

A. “repeatedly substantially simultaneously activating”

Each of the asserted claims of TDS’s '481 and '561 patents includes the limitation, “repeatedly substantially simultaneously activating.” ^[1] The district court construed this limitation as follows:

The term repeatedly means “repeating” in its ordinary sense, and that the repetitions be fast enough such that the composite color is actually perceived by the viewer. The term substantially, simultaneously activating means that during some portion of this period (defined as repeatedly), the two separate lights are on at the same time.

Tex. Digital Sys. Inc. v. Telegenix, Inc., No. 3:98-CV-1537-BF, slip op. at 11, 2000 WL 1801849 (N.D.Tex. Dec. 6, 2000).

Telegenix argues that the district court erred by requiring merely that the lights be on simultaneously, instead of requiring that the activation of each light begin at substantially the same time. According to Telegenix, the district court improperly separated the adverbs “substantially simultaneously” from the verb it modifies, “activating,” and thereby failed to require that the light emitting diodes (“LEDs”) must be activated, or turned on, at the same time.

According to TDS, the crucial word in the phrase is “repeatedly,” which would signal to one of skill in the art that the invention activates fight sources repeatedly within the “refreshing period” or “repetition period” "within which humans do not detect pulses due to the principle of “persistence of vision.” TDS argues that because one of skill in the art would understand that the claim re[*1206] fers to simultaneously activating light sources of different colors at some time during that critical repetition/refreshing period, and because humans cannot detect changes, pulses, or activations that occur within that period, the question of whether one begins activation of the LEDs at the exact same time or whether one simply ensures that both of the LEDs are on at some time during the repetition period is irrelevant. Thus, according to TDS, one of skill in the art would not interpret the claim term “repeatedly substantially simultaneously activating the light sources” to limit the starting point of the LED activations.

The district court correctly construed the term “repeatedly.” However, the district court’s construction of the overall phrase “repeatedly substantially simultaneously activating” was in error and ignored the meaning of the term “activating.” We begin by ascertaining the ordinary meaning to one skilled in the art. See Specialty Composites, 845 F.2d at 986, 6 USPQ2d at 1604. According to a relevant technical dictionary, to activate is “[t]o start an operation, usually by application of an appropriate enabling signal.” Modern Dictionary of Electronics 20 (6th ed.1984). We presume that the word used in a claim carries this ordinary meaning, but this presumption may be rebutted. See CCS Fitness, 288 F.3d at 1366, 62 USPQ2d at 1662. Here, the intrinsic evidence is entirely consistent with the dictionary definition, and there is nothing in the record to suggest that “activating” means other than what its dictionary definition would suggest, i.e., starting the operation or turning on. We conclude that the presumption has not been rebutted, and thus the ordinary meaning controls.

TDS has argued that “activating” can mean “being on.” Certainly, once activated, a lamp might accurately be described as “being on.” But the claim does not refer to the state of the lamps as being “substantially simultaneously activated.” The words used, which serve as the focus of the claim construction analysis, call for “substantially simultaneously activating” the lamps, and the ordinary meaning of that phrase requires that during some portion of the period defined as “repeatedly,” the two separate lights are turned on at the same or nearly the same time.

B. “selectively controlling the durations of the time intervals of activation”

Claims 1 and 3 of both the '481 and '561 patents include the limitation “selectively controlling the durations of time intervals of activation.” In its Markman ruling, the district court explicitly refused to provide a distinct definition for this limitation, deeming the phrase “sufficiently defined.”

Telegenix contended before the district court, and reiterates on appeal, that this limitation means “specifically controlling the length of time that individual pulses are activated to vary the amount of light emitted from a light source.” Telegenix argues that by his disclosure in the specification, the inventor limited the claims to varying color using pulse width modulation (“PWM”), a technique that varies the duration of individual pulses. Telegenix further contends that the inventor limited the asserted claims to PWM by statements and amendments during prosecution of the patents in suit.

TDS responds that the claims are not limited to the particular PWM technique suggested by Telegenix. TDS urges that the claim language uses the plural form of both “durations” and “time intervals” and thus is consistent with an interpretation in[*1207] which color is controlled with “more than one pulse and includes multiple activations of the same LED within the repetition period.” In other words, TDS urges a claim construction that would cover devices which change perceived light intensity by varying either the width of the pulses or the number of pulses.

The words of the claim require “controlling the durations” of the “time intervals of activation.” The plain meaning of “controlling the durations” indicates that the claimed invention requires variation of the duration of individual time intervals, or controlling the width of pulses, during which the LEDs are activated, e.g., PWM. This plain meaning is consistent with the specification of the '481 and '561 patents. The structures shown in Figures 9 and 11 of the '481 patent and Figure 1 of the '561 patent depict circuitry for driving the LEDs using PWM. As shown in Figure 9 of the '481 patent, the circuitry includes at least one counter 71f for each color connected to a corresponding memory 76 containing data regarding the amount of primary color activation required to produce the desired color. '481 patent, col. 4, 11. 24-59. The counter and corresponding memory are connected to a “flip-flop” 73 which provides the appropriate output to generate the desired color. Id. According to the specification, the “output of the flip flop 73 will be at a high logic level for a period of time proportional to the data” loaded into the counter 71f from the memory 76. Id. at col. 4, 11. 51-52. Thus, the circuitry controls color by setting the output “at a high logic level for a period of time proportional” to the desired color data.

Moreover, the prosecution history is consistent with this interpretation of the claim language. During prosecution of the '481 patent, the inventor distinguished pri- or art on the basis of PWM. The Patent Office initially rejected all claims in the application as obvious in view of the Kaelin reference, which taught that “LED color display elements can be varied by applying variable timed pulses to the individual diodes.” The applicant responded by submitting new claims and arguing that the invention “controls] the durations of the pulses that are applied to the primary color light sources in the selected display areas to control the portions of the primary color light signals, to thereby control the color of the exhibited display unit.”

On the basis of our review of the ordinary meaning of the words themselves, we conclude that this limitation requires control of pulse width. This is entirely consistent with the intrinsic record. Contrary to TDS’s argument, introducing multiple pulses of identical duration during the repetition period does not effect control of pulse duration. Where multiple pulses of identical duration are introduced during a single repetition period, pulse duration remains constant and color is controlled not by varying “the durations of the time intervals of activation” of pulses, as called for in the claims themselves, but by varying the number of constant duration pulses applied. Such a technique does not set the output “at a high logic level for a period of time proportional” to the desired color data nor does it “control the durations of the pulses,” and thus is inconsistent with the specification and prosecution history.

We conclude that “selectively controlling the durations of the time intervals of activation” means controlling the width of pulses during repetition periods.

C. “color control means”

The “color control means” limitation appears in claims 2, 4, and 7 of the '481 patent, and claims 2 and 4 of the '561 patent. The limitation appearing in claim 2 of the '481 patent is representative:

[*1208] color control means for selectively controlling the durations of the pulses applied to the light sources in the selected display areas to control the portions of the primary color light signals emitted therefrom, to thereby control the color of the exhibited display unit.

’481 patent, col. 9, 11. 59-64. The district court construed this limitation to be a means-plus-function limitation — a conclusion with which we agree. Neither party disputes that this limitation is subject to 35 U.S.C. § 112, paragraph six:. That paragraph states:

An element in a claim for a combination may be expressed as a means or 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.

35 U.S.C. § 112, para. 6 (2000).

“Because this limitation is expressed in ‘means plus function’ language and because it does not recite definite structure in support of its function, it is subject to the requirements of 35 U.S.C. § 112, ¶ 6 (1994).” B. Braun Med., Inc. v. Abbott Labs., 124 F.3d 1419, 1424, 43 USPQ2d 1896, 1899 (Fed.Cir.1997). The first step in construing such a limitation is to identify the function of the means-plus-function limitation. Micro Chem,., Inc. v. Great Plains Chem. Co., 194 F.3d 1250, 1258, 52 USPQ2d 1258, 1263 (Fed.Cir.1999). The next step is to identify the corresponding structure in the written description necessary to perform that function. Id. “Structure disclosed in the specification is ‘corresponding’ structure only if the specification or prosecution history clearly links or associates that structure to the function recited in the claim.” Braun, 124 F.3d at 1424, 43 USPQ2d at 1900.

The district court instructed the jury concerning the claimed function:

The color control means performs the function of selectively controlling the on times of the light sources to control the portions of primary color light signals for controlling the color of the composite light signal.

The district court described the corresponding structure from the specification as “the structure that performs as disclosed in the specification of the display decoder and decoder driver.”

Telegenix argues that the district court erred by misidentifying both the claimed function and the corresponding structure from the specification. We agree. The function recited in the asserted claims does not include “selectively controlling the on times of the light sources.” Instead, the claim recites “selectively controlling the durations of the pulses applied to the light sources.... ” To the extent that the district court failed to follow the claim language in defining the function, it erred. See Micro Chem., 194 F.3d at 1258, 52 USPQ2d at 1263 (“The statute does not permit limitation of a means-plus-function claim by adopting a function different from that explicitly recited in the claim.”).

The district court further erred in its identification of “the display decoder and decoder driver” as the corresponding structure in the specification.

Section 112, ¶ 6, as is well-documented, was intended to permit use of means expressions without recitation of all the possible means that might be used in a claimed apparatus.... The price that must be paid for use of that convenience is limitation of the claim to the means specified in the written description and equivalents thereof.

O.I. Corp. v. Tekmar Co., 115 F.3d 1576, 1583, 42 USPQ2d 1777, 1782 (Fed.Cir.1997) (citations omitted). The duty to link[*1209] or associate structure in the specification to the recited function is the quid pro quo for the convenience of employing § 112, paragraph 6. Braun, 124 F.3d at 1424, 48 USPQ2d at 1899. In the specification, the structure linked to the recited function of “selectively controlling the durations of the pulses applied to the light sources” includes the memory and counter circuitry illustrated in Figure 9. See '481 patent, col. 4, 11. 24-59. It was error for the district court to omit this structure from its claim construction of the color control means.

Accordingly, we conclude that the “color control means” performs the function of selectively controlling the durations of the pulses applied to the light sources to control the portions of the primary color light signals, to thereby control the color of the exhibited display unit. Moreover, we hold that the corresponding structure includes the memory 76, the counters 71e and 71f, the flip-flop 73, and associated connection circuitry illustrated in Figures 5 and 9. The color control means is limited to this corresponding structure and equivalents thereof.

D. “display areas” and “background area”

The patents in suit recite “display areas” and “background area” at several locations in the asserted claims. For example, claim 1 of the '619 patent recites “a plurality of variable color display areas ...; [and] a variable color background area....” The district court instructed the jury:

Display areas and the background areas “include any illuminated pixel anywhere on the display device with background pixels illuminated to substantially surround the illuminated display area pixels.” As the image for illuminated display area changes, so does the adjacent illuminated background area.”

Telegenix argues that the jury should have been instructed that the display areas are distinct from the background areas, that display areas cannot become background areas, and that background areas cannot become display areas. TDS argues that the claims encompass display areas arranged in the form of an array or matrix of areas, and as such the display areas and background areas are interchangeable.

Beginning with the words of the claims themselves, the dictionary meaning of display is “[a] visually observable presentation of information.... ” Illustrated Dictionary of Electronics 147 (3rd ed.1985). Background is defined as: “[the] context or supporting area of a picture....” Id. at 43. Thus, the ordinary meaning of “display area,” as reflected in these dictionary definitions, is an area designated to portray information. Background is ordinarily understood to provide the context or contrasting reference against which the displayed information is presented. The ordinary meaning of these limitations does not indicate that the display and background areas are interchangeable.

The specification of the '619 patent is consistent with an interpretation in which the display and background areas are distinct and not interchangeable. For example, the written description describes the invention as including “a variable color display area” and “a variable color background area 32, substantially surrounding the display area.” '619 patent, col. 2, 11. 16-21. The specifications of the patents in suit do not establish that display areas can become background areas, nor do they allow for their interchangeable use.

The specification of the '890 patent describes illuminating selected display areas in the background color to “blend with the background to provide maximum color contrast.” '890 patent, col. 2, 11. 41-54. However, this establishes only that the[*1210] inventor contemplated that display areas could junction similar to the background areas, not that the display and background areas could be interchangeable.

Looking to the prosecution history, there is additional evidence supporting a construction that the display and background areas are mutually exclusive. The inventor stated in response to a rejection: “[claims 1 and 2], similar to claim 13 which was not explicitly rejected, are distinguished from the prior art by the recitation of background regions separated from the display areas by opaque walls. No reference of the record describes explicitly defined background regions.” This evidence of manifest exclusion or restriction represents a clear disavowal of claim scope. See Teleflex, 299 F.3d at 1325, 63 USPQ2d at 1381. In doing so, the paten-tee expressly limited background areas to explicitly defined regions. In addition, the patent examiner stated in his notice of allowance, “[t]he prior art does not show the combination of variable color display areas and a variable color background area; these being two discrete, distinct components of the device. It is this distinction which, as claimed, is deemed allowable over the prior art.”

The ordinary meaning of the words of the claims, coupled with the patentee’s statements in the specification and during prosecution, establish that the district court’s instruction was in error. Although the specification indicates that a display area can be illuminated in the background color to “blend with the background to provide maximum color contrast,” there is no corresponding indication that the background areas can be illuminated in the display color. The district court’s construction that background areas can include “any illuminated pixel anywhere on the display device” is incompatible with the patentee’s statements during prosecution expressly limiting the background areas to “explicitly defined background regions.”

Moreover, if the background and display areas could each include “any illuminated pixel,” the background area would not be different in nature or quality from the display area. Such a proposition is inconsistent with the language of the claims, in which the inventor claimed a device having two types .of areas, and with the specification which describes distinct display areas and background areas.

On the basis of the ordinary meaning of the words of the claim and the intrinsic evidence, we conclude that these limitations should be construed as follows: display areas include any illuminated pixel anywhere on the display device, other than background area pixels in defined background regions. The background area pixels substantially surround the illuminated display area pixels. Display area pixels may be illuminated in the background col- or, but background area pixels may not be illuminated in the display color.

E. “display areas arranged in a pattern”

The limitation “display areas arranged in a pattern” appears in the asserted claims of the '481 and '619 patents, as well as claim 4 of the '890 patent. The district court construed “pattern” to mean “having a systematic arrangement.” Tele-genix argues that this construction is too general, and the limitation should be limited to a seven-segment display pattern, for example, that shown in Figures la-c of the '890 patent. TDS responds that this limitation is not limited to a seven-segment display or any other fixed pattern, and that the scope of the claims is broad enough to encompass a matrix display.

Where “pattern” is described in the specifications of the patents in suit, the seven-segment display is listed as an ex[*1211] ample of the preferred font. The '481 patent specification describes “seven elongated display segments a, b, c, d, e, f, g, arranged in a conventional pattern.” '481 patent, col. 2, 11. 24-25. The preferred embodiment of the '890 patent is described as including “a variable color display area consisting of seven segments 31 arranged in a well known 7-segment font.” '890 patent, col. 2, 11. 16-18. Nowhere in the specification is the limitation “display areas arranged in a pattern” restricted, explicitly or implicitly, to the seven-segment arrangement of the preferred embodiment.

Telegenix does not dispute that the patents in suit describe the seven-segment pattern in exemplary language. Instead, Telegenix argues that U.S. Patent No. 4,086,514 (“'514 patent”) establishes that the same inventor represented matrix displays and seven-segment displays as two separate embodiments of the same invention. We fail to understand the relevance of Telegenix’s argument. Whether or not the claims in an unrelated patent are broad enough to encompass both a matrix and the familiar seven-segment pattern, this proposition sheds no fight on whether the claims of the patents in suit are limited to the seven-segment pattern. See Abbott Labs. v. Dey, L.P., 287 F.3d 1097, 1104, 62 USPQ2d 1545, 1550 (Fed.Cir.2002) (finding the relationship between two unrelated patents, although having common subject matter, a common inventor, and the same assignee, “insufficient to render particular arguments made during prosecution of [one of the patents] equally applicable to the claims of [the other patent]”).

Referring to the prosecution history, the Examiner’s Statement of Reasons for Allowance for the '619 patent stated, “In this manner, multicolored arrays (i.e., col- or cathode ray tube displays such as Take-da, of record) in which there is no physical distinction between a foreground or background pixel (display area), are distinguished from by the claimed subject matter.” Although the prosecution history may help define the scope of a term if relevant, see Southwall Techs., Inc. v. Cardinal IG Co., 54 F.3d 1570, 1576, 34 USPQ2d 1673, 1676 (Fed.Cir.1995), this Examiner’s statement has no bearing on the meaning of the term “pattern.” Nor does this statement limit the scope of “pattern” to the familiar seven-segment font.

Accordingly, because there is nothing in the claims or the intrinsic evidence of record to indicate otherwise, we conclude that this limitation was correctly construed by the district court according to the ordinary meaning of “pattern.” The district court’s construction, “having a systematic arrangement,” is not in error.

F. “means for selectively activating said display fight sources”

The limitation “means for selectively activating said display fight sources” appears in claim 1 of the '619 patent. The district court instructed the jury:

The function of this means-plus function element is to control the activation of fight sources to illuminate the display area in a first color and the background area in a second color, different from the display area’s first color. The function is to activate the display area by passing current through selected fight sources of the display area and the background area. The structural components are specified in Figures 3 and 4.

Telegenix argues that the court’s construction is too broad, and misled the jury to believe that this limitation can be met by any structure broadly suggested by Figure 3, which shows only a block diagram. Telegenix requested an instruction limiting the structure to the circuitry shown in Figure 4. Telegenix also argues[*1212] that “passing current through selected light sources” incorrectly identifies the function.

TDS argues that its expert testified that “means for selectively activating” includes hardware, software, and/or firmware for passing current through selected light sources, as supported by Figures 3 and 4. TDS argues that a block diagram such as that shown in Figure 3 may describe structure.

TDS argues in favor of a broad interpretation of this claim limitation in reliance on the testimony of its expert. “[E]xtrinsie evidence in general, and expert testimony in particular, may be used only to help the court come to the proper understanding of the claims; it may not be used to vary or contradict the claim language.” Vitronics, 90 F.3d at 1584, 39 USPQ2d at 1578. Where the patent documents are unambiguous, expert testimony regarding the meaning of a claim is entitled to no weight. Id. “ ‘Any other rule would be unfair to competitors who must be able to rely on the patent documents themselves, without consideration of expert opinion that then does not even exist, in ascertaining the scope of a patentee’s right to exclude.’ ” Id. (quoting Southwall Techs., 54 F.3d at 1578, 34 USPQ2d at 1678-79). Thus, where the patent documents are unambiguous, reliance upon the testimony of TDS’s expert witness would be improper.

We hold, and the parties do not dispute, that this limitation is in means-plus-function form and thus is subject to 35 U.S.C. § 112, paragraph six. As stated above, in construing such a limitation, the task of the district court is first to identify the function recited for the limitation and next to identify the corresponding structure in the written description necessary to perform that function. See Micro Chem., 194 F.3d at 1258, 52 USPQ2d at 1263.

The district court misidentified the recited function by including in the construction, “The function is to activate the display area by passing current through selected light sources of the display area and the background area.” This reference to “passing current” has no basis in the claim language. See Generation II Orthotics, 263 F.3d at 1364-65, 59 USPQ2d at 1926 (“When construing the functional statement in a means-plus-function limitation, we must take great care not to impermissibly limit the function by adopting a function different from that explicitly recited in the claim.”); Micro Chem., 194 F.3d at 1258, 52 USPQ2d at 1263 (“The statute does not permit limitation of a means-plus-function claim by adopting a function different from that explicitly recited in the claim.”).

Likewise, the district court’s identification of the corresponding structure was incomplete. The description in the specification of the structure corresponding to the recited function is not limited to Figures 3 and 4, as instructed, but also includes the written description accompanying these Figures. See '619 patent, col. 3, 11. 34-68, and col. 4,11. 1-61. Moreover, as Figure 3 and its accompanying text serve merely as overview for introducing and explaining Figure 4, the corresponding structures must necessarily be found in Figure 4.

We conclude that the “means for selectively activating” performs the function of “illuminat[ing] certain of said display areas in a first color, and said background light sources, to illuminate said background regions in a second color different from said first color.” The corresponding structure in the specification is described in Figure 4 and the accompanying written description, including the overview provided by Figure 3 and the written description accompany[*1213] ing that Figure. The “means for selectively activating” is limited to this corresponding structure and equivalents thereof.

G. “converter means”

The phrase “converter means” is a limitation recited in claim 4 of the '890 patent. The district court interpreted this phrase to mean:

The converter means includes firmware, software and/or hardware that functions to convert said display color control signals to obtain complementary color control signals.

Telegenix argues that this interpretation is unsupported, because no software or firmware is mentioned anywhere in the specification. Telegenix argues that the structure disclosed for the “converter means” is the multiplexer and inverter arrangement described in the written description and figures.

TDS argues that its expert testified that one of ordinary skill in the art would have appreciated that the converter means could be implemented in hardware, software, and/or firmware. TDS argues that the function of the converter means includes providing a complementary color in response to the selected display area color, and the multiplexer does not perform this function. Instead, according to TDS, only the inverter performs this function.

There is no dispute that “converter means” is a means-plus-function limitation within the meaning of section 112, paragraph 6. Again, in construing such a limitation, the task of the district court is to first identify the function recited for the limitation, and next to identify the corresponding structure in the written description necessary to perform that function. See Micro Chem., 194 F.3d at 1258, 52 USPQ2d at 1263.

In its construction of the “converter means,” the district court failed to identify the corresponding structure from the specification. The district court correctly performed the first step by identifying the claimed function, “to convert said display control color signals to obtain complementary color control signals.” However, the court’s description of the corresponding structure as “including] firmware, software and/or hardware” has no basis in the specification. TDS essentially admits in its brief that the testimony of TDS’s expert was used to broaden the court’s view of the corresponding structure beyond that disclosed in the specification and prosecution history.

TDS argues that the structure identified by the district court could qualify as equivalent structure under section 112, paragraph six. This argument misunderstands the deficiency in the district court’s construction. Even if software and firmware could be equivalent structures under section 112, paragraph six, the court failed to correctly perform the second step of identifying the structure disclosed in the written description as corresponding to the recited function. Instead, the court identified a broad array of possible structures not mentioned anywhere in the specification.

We conclude that the district court erred in construing this limitation. Where the patent documents are unambiguous, expert testimony regarding the meaning of a claim is entitled to no weight. Vitronics, 90 F.3d at 1584, 39 USPQ2d at 1578. The “converter means” performs the function of converting the display color control signals to obtain complementary color control signals. The corresponding structure includes inverters 26a-c depicted in Fig. 3 and described in the specification at col. 3, 11. 31-39, and col. 4,11. 8-27. The “convert[*1214] er means” is limited to this corresponding structure and equivalents thereof.

H. “first means” and “second means” for carrying color control signals

The “first means” and “second means” are limitations recited in claim 4 of the '890 patent. The district court instructed the jury:

The first means for carrying includes functions that are performed by electrical paths which are non-inverting buses for the red and green LEDs shown in Figure 3 and described in Column 3, lines 23-30. First means includes “firmware, software and/or hardware that function to carry the information which determines the display area (character) color.”

The court further instructed:

The second means for carrying includes functions that are performed by electrical paths which are inverting buses for the red and green LEDs shown in Figure 3 and described in column 3, lines 31-35. Second means for carrying includes “any firmware, software and/or hardware that function to carry complementary control signals.”

Telegenic repeats its arguments that the specification does not disclose “firmware, software, and/or hardware” that performs the claimed functions. Telegenix argues that the structure should be limited to inverting and non-inverting electrical buses.

TDS argues that Figure 2 shows structure for carrying the color control signals in the form of signal lines connecting the Display Color Control block 21 and the Complement Color Control block 22 with the Variable Color Display block 11.

Again, the parties do not dispute that the “first means” and “second means” are in means-plus-function form and thus are subject to 35 U.S.C. § 112, paragraph six.

The district court misidentified both the recited function and the corresponding structure with respect to “first means” and “second means.” Instead of identifying the function recited for the first means, the district court’s instruction to the jury indicated that the function was “to carry the information which determines the display area (character) color.” We disagree. This language appears nowhere in claim 4, and unnecessarily limits the function actually recited in claim 4, “carrying selective display color control signals.” See Generation II Orthotics, 263 F.3d at 1364-65, 59 USPQ2d at 1926 (‘When construing the functional statement in a means-plus-function limitation, we must take great care not to impermissibly limit the function by adopting a function different from that explicitly recited in the claim.”). Likewise, the district court erred in identifying the recited function of the second means. The recited function is “for carrying said complementary color control signals.”

Concerning the district court’s identification of corresponding structure, it is undisputed that such structure includes the non-inverting buses described in the specification. However, the district court ventures beyond the specification to include in its construction “any firmware, software and/or hardware” that performs the identified function. Committing the same error as with the “converter means,” the district court relied on expert testimony to broaden its interpretation of the corresponding structure beyond that appearing in the specification.

Although TDS argues that Figure 2 identifies structure broader than the inverting and non-inverting buses described in the written description, Figure 2 fails to describe any structure for the first and second means sufficient to comport with[*1215] section 112, paragraph six. If a patentee fails to disclose an adequate corresponding structure in the specification, the patentee may fail to satisfy the bargain embodied in the statutory quid pro quo of section 112, paragraph six. See Kemco Sales, Inc. v. Control Papers Co., 208 F.3d 1352, 1360, 54 USPQ2d 1308, 1313 (Fed.Cir.2000). Notwithstanding its adequacy, Figure 2 provides no support whatsoever for the district court’s identification of the corresponding structure as including “any firmware, software and/or hardware.”

We conclude that the district court’s claim construction for “first means” and “second means” was flawed to the extent that it misidentified the functions recited in claim 4 of the '890 patent and included in the corresponding structure “any firmware, software and/or hardware.” See Vitronics, 90 F.3d at 1585, 39 USPQ2d at 1579 (“Because the specification clearly and unambiguously defined the disputed term in the claim, reliance on this extrinsic evidence was unnecessary and, hence, legally incorrect.”). The recited function of the first means is “carrying selective display color control signals,” and the corresponding structure includes “electrical paths which are non-inverting buses for the red and green LEDs shown in Figure 3 and described in Column 3, lines 23-30.” The recited function of the second means is “carrying said complementary color control signals,” and the corresponding structure includes “electrical paths which are inverting buses for the red and green LEDs shown in Figure 3 and described in column 3, lines 31-35.” The first means and second means are limited to the identified corresponding structure and equivalents thereof.

I. “control means for selectively coupling said light sources”

The “control means for selectively coupling” appears in claim 4 of the '890 patent. Claim 4 of the '890 patent recites:

control means for selectively coupling said light sources in said display areas to said first means, for causing selective ones of said display areas to illuminate in a selected color defined by said display color control signals, and to said second means, for causing the remaining display areas to illuminate in a substantially complementary color defined by said complementary color control signals.

’890 patent, col. 10,11.16-23.

The district court’s instruction to the jury construing “control means” was:

Control means includes any firmware, software, and/or hardware that functions to selectively couple the light sources in the display areas to said first means for carrying thereby causing the selective ones of the display areas to illuminate in a selected color.... Control means is defined as a multiplexer. Multiplexers serve to selectively couple each display area of a display device to non-inverting and inverting buses in order to illuminate the display areas with either the desired color or a substantially complimentary color in accordance with the output of the decoder. The decoder output is respectively coupled to the display areas. The multiplexer simultaneously couples the display areas to the display control bus and couples the converted display signal to the background areas of the display device.

Telegenix argues that inclusion of “any firmware, software, and/or hardware” was error. Telegenix argues that the Statement of Reasons for Allowance in the prosecution history of the '890 patent shows that the inventor limited the claims to require a hardware multiplexer, thus firmware or software multiplexers would be excluded.

[*1216] TDS concedes that the “control means” must include a multiplexer, but TDS contends that the circuit shown in Figure 4 of the '890 patent is not the only implementation of a multiplexer, again citing expert testimony in support.

The limitation “control means” is in means plus function form, and neither party disputes the district court’s identification of the recited function, which we conclude is correct.

However, for the same reasons announced earlier with regard to the “converter means,” the district court erred by including “any firmware, software, and/or hardware” in its identification of the corresponding structure. See Vitronics, 90 F.3d at 1585, 39 USPQ2d at 1579 (“Because the specification clearly and unambiguously defined the disputed term in the claim, reliance on this extrinsic evidence was unnecessary and, hence, legally incorrect.”). We can find no support in the specification or prosecution history for such a broad array of structures. Instead, the specification describes a hardware multiplexer at col. 5,11. 45-68 and col. 6,11. 1-20, illustrated in Figures 3 and 4. We conclude that the correct construction of “control means” is that of the district court with the phrase “includes any firmware, software, and/or hardware that” excised from the first sentence and the phrase “a multiplexer” at the end of the second sentence replaced with the phrase — the multiplexer shown in Figure 4 and described in the accompanying written description, and equivalents thereof — .

II. Prejudicial Error

Telegenix has shown that the district court erred in construing limitations of the claims, but this alone is not enough to challenge jury instructions with respect thereto — the standard of review for jury instructions is prejudicial legal error. See Jamesbury, 756 F.2d at 1558, 225 USPQ at 255. Thus, to prevail, the party challenging a jury instruction “must demonstrate both that the jury instructions actually given were fatally flawed and that the requested instruction was proper and could have corrected the flaw.” Biodex, 946 F.2d at 862, 20 USPQ2d at 1261; accord Ecolab, 285 F.3d at 1372-73, 62 USPQ2d at 1356-1357.

Although TDS argues that Telegenix has failed to demonstrate prejudice from these claim construction errors, Telegenix correctly points to the record, which shows that Telegenix warned the district court concerning the court’s failure to properly construe the means-plus-funetion limitations and proposed constructions of the disputed claim limitations that would have corrected the flaws.

On this record, we conclude that the claim construction errors committed by the district court were prejudicial. Accordingly, we vacate the decision of the district court and remand for a new trial of both liability and damages. To assist the district court on remand, we address the allegations of error regarding the admissibility of the challenged testimony of Brent W. Brown (“Brown”) and J. Carl Cooper (“Cooper”), and the district court’s reliance on the Supreme Court’s decision in Wine Railway Appliance Co. v. Enterprise Railway Equipment Co., 297 U.S. 387, 56 S.Ct. 528, 80 L.Ed. 736 (1936).

III. Admissibility of Evidence

We review a trial court’s decision to exclude evidence for abuse of discretion. Beech Aircraft Corp. v. Rainey, 488 U.S. 153, 172, 109 S.Ct. 439, 102 L.Ed.2d 445 (1988). To be admissible, expert testimony must “assist the trier of fact to understand the evidence or to determine a fact in issue.” Fed.R.Evid. 702; Kumho Tire [*1217] Co. v. Carmichael, 526 U.S. 137, 147, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999).

Telegenix argues that the district court abused its discretion by excluding the evidence offered by Brown, and by admitting the testimony of Cooper respecting damages. We first address Telegenix’s argument with respect to Brown and then with respect to Cooper.

A. Brown

Telegenix argues that the district court erroneously excluded the testimony of Brown, an engineer who developed a variable color LED display in the early 1980’s. Telegenix argues that Brown’s testimony, along with his 1982 patent application, would have shown that the patents in suit were in public use prior to one year before the original application in 1986.

The district court refused to admit Brown’s testimony, stating:

He testified that he had a prototype of a multicolor display unit somewhere prior to 1985. I think he testified '83 or so, somewhere around there, but he didn’t testify that it was out in the public anywhere or that any were sold prior to 1986. He just couldn’t remember. That’s the type of unreliable evidence that is difficult to make a determination whether or not to admit to a jury.

The district court also stated:

It’s too dangerous to submit this evidence to the jury based upon the testimony of Mr. Brown and their offer of proof because Mr. Brown is just uncertain of the facts and circumstances surrounding when the invention actually got out into the public.... The Court finds that the uncorroborated testimony of Mr. Brown would be confusing to the jury....

Telegenix argues that the district court erred by excluding Mr. Brown’s testimony as lacking in sufficient corroboration. Telegenix contends that because his evidence was not offered as a party seeking to prove priority, no corroboration is required. Telegenix’s argument misreads our caselaw on corroboration. “[C]orrobo-ration is required of any witness whose testimony alone is asserted to invalidate a patent, regardless of his or her level of interest.” Finnigan Corp. v. Int’l Trade Comm’n, 180 F.3d 1354, 1369, 51 USPQ2d 1001, 1012 (Fed.Cir.1999). Thus, the district court correctly required corroboration for Brown’s testimony.

Telegenix argues that, even if Brown’s testimony required corroboration, his 1982 patent application and other documentary and physical evidence provided sufficient corroboration for his testimony. Telegenix argues that the court erred by excluding the 1982 unissued patent application as corroborating evidence, citing Sandt Technology v. Resco, 264 F.3d 1344, 1351, 60 USPQ2d 1091, 1094 (Fed.Cir.2001).

We assess corroboration according to the factors enumerated in Woodland Trust v. Flowertree Nursery, Inc.:

(1) the relationship between the corroborating witness and the alleged prior user, (2) the time period between the event and trial, (3) the interest of the corroborating witness in the subject matter in suit, (4) contradiction or impeachment of the witness’ testimony, (5) the extent and details of the corroborating testimony, (6) the witness’ familiarity with the subject matter of the patented invention and the prior use, (7) probability that a prior use could occur considering the state of the art at the time, and (8) impact of the invention on the industry, and the commercial value of its practice.

[*1218] 148 F.3d 1368, 1371, 47 USPQ2d 1363, 1366 (Fed.Cir.1998). “Documentary or physical evidence that is made contemporaneously with the inventive process provides the most reliable proof that the inventor’s testimony has been corroborated.” Sandt, 264 F.3d at 1350-51, 60 USPQ2d at 1094.

Despite Telegenix’s argument, the district court did not refuse to consider Brown’s 1982 unissued patent application, but instead expressly considered it for corroboration purposes. Judge Stickney stated in open court, “Now, Mr. Brown’s testimony is not corroborated other than by his patent application, which the Court finds is insufficient corroboration.”

However, the district court did refuse to consider Brown’s '114 patent and physical evidence for corroboration purposes because it was not prior or contemporaneous evidence, as Telegenix concedes. TDS argues that the physical evidence offered with Brown’s testimony was properly excluded because it was built after the effective date of the patents in suit. Whether or not the district court erred in refusing to consider this evidence for corroboration purposes, Telegenix faces a particularly high hurdle in attempting to demonstrate abuse of discretion in light of the stringent standard for corroboration. See Juicy Whip, Inc. v. Orange Bang, Inc., 292 F.3d 728, 741-43, 63 USPQ2d 1251, 1260-61 (Fed.Cir.2002). In the absence of further contemporaneous corroborating evidence, we are unable to conclude that the district court abused its discretion in refusing to admit Brown’s testimony for lack of corroboration.

The district court excluded Brown’s testimony for the further reason that Brown gave uncertain testimony concerning the date of public use: “Mr. Brown is just uncertain of the facts and circumstances surrounding when the invention actually got out into the public ... he just really doesn’t remember anything.” The district court cited Federal Rule of Evidence 403 (“FRE”) and found that Brown’s testimony would be confusing to the jury.

Although the record before us indicates that Brown gave clear and definite testimony concerning certain facts related to public use, it is also clear that Brown could not recall the details:

Q: You said you sold the company — did you sell multicolor displays using red and green LEDs prior to your sale of ISE to Bray in 1983?

A: I can’t honestly remember at that point.

Q: Did Bray sell those devices after you sold the company to Bray and moved over to that company?

A: Yes. That was a product line that we continued to develop and was being sold when I bought the company back in 1986, in December.

Q: They were sold prior to 1986?

A: Yes.

Q: For approximately how long?

A: I can’t tie it down exactly because the — it was an ongoing development, and I was running both companies, the whole division. So I can’t tie it any closer than that.

The district court ultimately excluded the evidence because Brown could not establish a particular date of public use, a critical consideration for a statutory bar. The court found Brown’s testimony unreliable and potentially confusing to the jury, and rightfully excluded it pursuant to FRE 403. On the basis of the record before us, we cannot conclude that the district court abused its discretion.

B. Cooper

Telegenix argues that the district court improperly admitted Cooper’s[*1219] revised expert report on patent damages. Telegenix contends that Cooper is unqualified to testify as an expert in this subject. Telegenix argues that Cooper’s report contained numerous flaws, including that Cooper’s profit calculations for Telegenix products were not based on the actual cost figures provided by Telegenix; that Cooper did not take into account that hypothetical licensing negotiations taking place in 1992 would have been with the inventor Mr. Havel, not with TDS; that Cooper did not account for revenues and profits from related products; and that Cooper’s premise that TDS did not license its patents was false because the evidence showed that another company took two licenses.

Despite Telegenix’s qualifications argument, the evidence supports a finding that Cooper was competent and qualified. Cooper owned and managed two patent licensing companies following his work in manufacturing displays in the early 1990’s.

The district court initially refused to admit Cooper’s second damages report because Cooper had revised the cost figures provided by Telegenix. The court ruled that Cooper must use the actual cost figures provided by Telegenix, and permitted Cooper to submit another report with the corrected figures. Although Telegenix argues that it did not receive the revised report until shortly before trial, due to the nature of the revisions Telegenix can hardly claim unfair surprise.

Telegenix’s other complaints allege no abuse of discretion by the district court. Instead, Telegenix takes issue with the content of Cooper’s opinion. As the district court stated, Telegenix’s complaints go to weight, not admissibility. We conclude that the district court did not abuse its discretion by admitting Cooper’s corrected expert report.

IV. Wine Railway

Telegenix argues that the district court erroneously relied on Wine Railway Appliance Co. v. Enterprise Railway Equipment Co., 297 U.S. 387, 56 S.Ct. 528, 80 L.Ed. 736 (1936), in sustaining the jury’s award of damages for acts of infringement dating back to 1992. Telegenix argues that it did not receive notice that it was infringing the patents in suit until 1998, and, by permitting liability for acts prior to 1998, the rule of Wine Railway undermines the notice requirement of 35 U.S.C. § 287.

TDS argues that the damages awarded by the jury were not limited by section 287 because TDS did not trigger operation of the statute. TDS contends that it could not have triggered operation of the statute because it did not produce or sell the patented product — there was no “failure so to mark” under 35 U.S.C. § 287(a). TDS argues that Wine Railway is still good law on which the district court properly relied.

The Supreme Court in Wine Railway held that the patent marking statute then in effect did not require a patentee who did not produce the patented device to give actual notice to an infringer before damages could be recovered. Although Wine Railway interpreted a predecessor to the current patent marking statute, we have applied Wine Railway to the modern statutory counterpart, 35 U.S.C. § 287. See Nike, Inc. v. Wal-Mart Stores, Inc., 138 F.3d 1437, 1443, 46 USPQ2d 1001, 1008 (Fed.Cir.1998); Am. Med. Sys., Inc. v. Med. Eng’g Corp., 6 F.3d 1523, 1538, 28 USPQ2d 1321, 1332 (Fed.Cir.1993); Bandag, Inc. v. Gerrard Tire Co., 704 F.2d 1578, 1581, 217 USPQ 977, 979 (Fed.Cir.1983).

Telegenix’s arguments reveal a misunderstanding of the patent marking statute. The statute does not specify when or under what circumstances damages may be recovered. Rather, it de[*1220] scribes circumstances that effect a forfeiture of damages:

In the event of failure so to mark, no damages shall be recovered by the pat-entee in any action for infringement, except on proof that the infringer was notified of the infringement and continued to infringe thereafter, in which event damages may be recovered only for infringement occurring after such notice.

35 U.S.C. § 287(a) (2000). Thus, section 287 “penalizes the use of unauthorized marks upon manufactured articles” and limits the extent to which damages may be recovered where products covered by a U.S. patent are sold without the notice defined in the statute. Wine Railway, 297 U.S. at 893, 56 S.Ct. 528. The recovery of damages is not limited where there is no failure to mark, i.e., where the proper patent notice appears on products or where there are no products to mark. Id. As the Supreme Court so aptly stated:

The idea of a tangible article proclaiming its own character runs through this and related provisions. Two kinds of notice are specified — one to the public by a visible mark, another by actual advice to the infringer. The second becomes necessary only when the first has not been given; and the first can only be given in connection with some fabricated article. Penalty for failure implies opportunity to perform.

Id. at 395. The district court did not err in its reliance on the rule of Wine Railway.

CONCLUSION

For the foregoing reasons, the decision of the district court is affirmed-in-part, reversed-in-part, and remanded.

AFFIRMED-IN-PART, REVERSED-IN-PART, and REMANDED.

COSTS

No costs.