Free Redacted Document - District Court of Delaware - Delaware

Case 1:04-cv-01371-JJF

Document 529

Filed 09/10/2007

Page 1 of 8

FISH & RICHARDSON P.C.
Suite
I1oo

9i9 N. Market Street P.O. Box 1114

Frederick P. Fish 1855-1930 W.K. Richardson 1859-1951

Wilmington, Delaware 19899-1114 Telephone 302 652-5070 Facsimile 302 652-0607

FILED UNDER SEAL August 31, 2007

REDACTED

Web Site www.fr.com William J. Marsden, Jr. 302 778-8401

The Honorable Joseph J. Farnan, Jr. United States District Court For the District of Delaware 844 King Street Wilmington, DE 19801
1%
ATLANTA AUSTIN BOSTON DALLAS DELAWARE NEW YORK SAN DIEGO SILICON VALLEY TWIN CITIES WASHINGTON, DC

Email [email protected]

Re:

Power Integrations' Motions in Limine Power Integrations, Inc. v. Fairchild Semiconductor International USDC-D. Del. - C.A. No. 04-1371 JJF

Dear Judge Farnan: Power Integrations submits this letter brief to present and attempt to resolve six discrete issues in anticipation of the trial set to begin in this action on September 17, 2007. Specifically, Power Integrations renews the motions filed earlier as D.I. 479, with the exception of previous motion number 4, which has been withdrawn, and raises two new motions.' Power Integrations hereby moves in limine to: (1) preclude any reference at trial to the reexamination of Power Integrations' circuit patents; (2) preclude Fairchild expert Dr. Paul Horowitz from relying on alleged "implied" findings by the jury in the first trial which contradict the Court's claim constructions; (3) obtain a more definite statement regarding Fairchild's invalidity contentions; (4) [withdrawn] (5) preclude any testimony from Gu-Yeon Wei, Fairchild's purported "expert on copying"; (6) preclude Fairchild expert Dr. Paul Horowitz from relying on alleged "prior art" not disclosed in his expert report and which lacks foundation; and (7) exclude "expert" testimony from fact witness James Beasom. The substance of motions in limine 1-3 and 5 remains unchanged, and Power Integrations has maintained the numbering of the motions from its previous brief in an attempt to prevent any confusion regarding the motions.

Case 1:04-cv-01371-JJF

Document 529

Filed 09/10/2007

Page 2 of 8

FISH & RICHARDSON P.C.

The Honorable Joseph J. Farnan, Jr. August 31, 2007

Page 2 1. FAIRCHILD SHOULD NOT BE ALLOWED TO INTRODUCE IRRELEVANT AND PREJUDICIAL EVIDENCE RELATED TO THE REEXAMINATION OF POWER INTEGRATIONS' PATENTS.

Fairchild has indicated that it intends to introduce its reexamination requests and the PTO's grant of reexam on three of the patents-in-suit during the upcoming trial. Those requests are irrelevant and unduly prejudicial to Power Integrations for several reasons, including that (1) different legal standards apply in the Patent Office, (2) more limited evidence is available in the Patent Office (due to the practical inability to file confidential information under seal there), and (3) there has been no substantive office action by the Patent Office in any of the reexam proceedings. As the Federal Circuit has noted, "the grant of a request for reexamination, although surely evidence that the criterion for reexamination has been met (i.e., that a `substantial new question of patentability' has been raised, 35 U.S.C. § 303), does not establish a likelihood of patent invalidity." Hoechst Celanese Corp. v. BP Chemicals Ltd., 78 F.3d 1575, 1584 (Fed. Cir. 1996). In light of this, courts faced with this exact question routinely exclude mention of reexamination to a jury. See Amphenol T & M Antennas Inc. v. Centurion Int'l Inc., 69 U.S.P.Q. 2d 1798, 1800 (N.D. 111. 2002) ("[T]elling the jury that the patent has been called into question by the Patent Office may significantly influence the jury's application of the presumption of validity and significantly prejudice [the patentee]. The prejudicial potential of this evidence far outweighs any probative value it might have. See Fed. R. Evid. 403."). In this case, Power Integrations strongly believes that the prior art raised in reexamination, and the prior art Fairchild will attempt to rely on at trial, does not come close to teaching the claimed inventions and, therefore, the reexamination will result, as do the majority of reexaminations, in the patentability of Power Integrations' claims being confirmed. Power Integrations has discussed this matter with Fairchild, but Fairchild maintains that it should be permitted to raise the issue at trial, and Fairchild has included three documents from the incomplete reexamination proceedings on its exhibit list (DX601, DX602, and DX603). Power Integrations therefore seeks an order under Federal Rules of Evidence 402 and 403 preventing Fairchild from offering evidence at trial, or raising in any manner, the reexamination of Power Integrations' patents or the standards governing reexamination, including but not limited to using any such materials allegedly as "impeachment" evidence or on cross examination. II. FAIRCHILD'S CIRCUIT EXPERT, DR. PAUL HOROWITZ, SHOULD NOT BE PERMITTED TO PROVIDE TESTIMONY BASED ON ANY IMPLIED JURY FINDINGS THAT CONTRADICT THE COURT'S CLAIM CONSTRUCTIONS.

Fairchild's expert on circuit technology, Dr. Paul Horowitz, testified in deposition that he will apply, in the validity trial, claim constructions different than those provided by the Court and different from what he himself advocated in the first trial.

Case 1:04-cv-01371-JJF

Document 529

Filed 09/10/2007

Page 3 of 8

FISH & RICHARDSON P.C.

The Honorable Joseph J. Farnan, Jr. August 31, 2007

Page 3
According to Dr. Horowitz's latest theory, the jury could not have reached their verdict of infringement using the Court's constructions. (See, e.g., Ex. A (Horowitz Tr. 1/30/07) at 200-205.)2 This issue is particularly significant with respect to the "soft start circuit" limitation in Power Integrations' '366 patent,3 where Dr. Horowitz apparently intends to testify that even though he previously acknowledged that the structures within the box labeled "soft start" in Figure 3 of the patent are required under the Court's means-plus-function construction of the term (Ex. B (Trial Tr.) D.I. 419 at 1086-88), those structures must not truly be required because the jury disagreed with Dr. Horowitz's ultimate conclusion and found that Fairchild infringed the patent. (Ex. A (Horowitz Tr. 1/30/07) at 203-04

Dr. Horowitz's new position presumes the jury construed the claims in a manner contrary to the Court's instruction, because the Court specifically directed the Jury to Figure 3 of the patent (among other things) for the corresponding structure for the "soft start circuit" limitation. (Ex. B (Trial Tr.) D.I. 420 at 1655.) As Power Integrations' expert Mr. Blauschild explained at trial, and consistent with Dr. Horowitz's own testimony, Figure 3 of the '366 patent shows that the soft start circuit (box 410) includes a latch (450), a comparator (460), and an AND gate (455). (Id., D.I. 417 at 489-90.)4 Dr. Horowitz's new argument, that those elements must not have been required to meet the soft start circuit limitation, runs contrary to the presumption that jurors follow instructions and apply them correctly. Dr. Horowitz's new position also ignores a legitimate (and in fact the only logical) explanation for the first verdict - that, consistent with the Court's claim construction, the jury agreed with both Dr. Horowitz and Mr. Blauschild that the soft start circuit includes the structures inside the "soft start" block in the figure, but then disagreed with Dr. Horowitz and agreed instead with Mr. Blauschild that the required structures were, in All exhibit citations refer to the accompanying declaration of William J. Marsden, Jr., unless otherwise noted. This issue is not limited to the '366 patent, however, because Power Integrations believes that Dr. Horowitz will also present inconsistent opinions regarding the '876 patent.
4

It was Dr. Horowitz who incorrectly asserted at the first trial that additional structures outside the soft start block, for example a low frequency oscillator, were also required by the claim. Fairchild now seeks to use that straw-man as its centerpiece of its own motion in limine because there was no such oscillator structure in its products - a fact never contested by Power Integrations. Fairchild should not be allowed to use the fact that the jury rejected Fairchild's own expert's clearly erroneous argument as a basis to assert that the jury ignored or misapplied the Court's construction.

Case 1:04-cv-01371-JJF

Document 529

Filed 09/10/2007

Page 4 of 8

FISH & RICHARDSON P.C.

The Honorable Joseph J. Farnan, Jr. August 31, 2007 Page 4 fact, present in the accused products. If Fairchild believes there was insufficient evidence for the jury to have so found, its proper course is a motion for a judgment as a matter of law, not to seek a tactical advantage in "applying" and perpetuating the perceived "error" in the validity proceeding. Any testimony from Dr. Horowitz based on a different "implied" or "presumed" claim construction would be improper and likely to confuse the jury in the second case or, worse yet, suggest to the second jury that it is free to disregard the Court's claim constructions in the manner Dr. Horowitz incorrectly asserts the first jury did. Power Integrations therefore seeks an order precluding Fairchild and its expert Dr. Horowitz from arguing or insinuating that the first jury must have made implied findings contrary to the Court's constructions. Fairchild and Dr. Horowitz are so unwilling to accept the first jury's rejection of their non-infringement positions that they have brought their own motion in limine on this same subject. (See D.I. 478.) Fairchild's motion is based on the fiction that Power Integrations "intends" to argue a "new" position "contrary" to the jury's finding of infringement. By its motion, Fairchild seeks free reign to allow its expert to change his opinions without concern that the new opinions will be impeached as inconsistent with his own prior sworn testimony. Power Integrations filed its opposition to Fairchild's motion and has addressed the issues more fully therein. (D.I. 485 at 4-8.) Nevertheless, because the issue is related to Power Integrations' motion, we note here that Power Integrations will present the same recitation of the claimed structure at the upcoming validity trial that its expert presented during the first trial, namely that the claimed soft start circuit includes the structures shown inside the "soft start" box in the Figure 3 of the '366 patent.5 III. THE COURT SHOULD PREVENT FAIRCHILD FROM BURYING ITS INTENDED PROOFS AMONG MORE THAN THREE DOZEN POTENTIAL INVALIDITY THEORIES ITS EXPERTS HAVE SET FORTH IN THEIR REPORTS.

Fairchild has buried the contentions it intends to present at trial among the more than 40 invalidity contentions set forth in Fairchild's expert reports. Fairchild's circuit 5 The Court will appreciate that the only structure specifically mentioned in Fairchild's motion is the low frequency oscillator, even though Fairchild makes a vague reference to "other structural components" several times. The real issue is not the low frequency oscillator. Power Integrations has never asserted, and the Court did not say, that the claim required any structures beyond the three shown inside the "soft start" block in the figures. Dr. Horowitz, on the other hand, needs to change his previous testimony to argue now that the soft start circuit does not require the specific structural components inside the soft start box, because those structures are not found in Fairchild's proposed prior art. Therefore, Fairchild's invocation of the low frequency oscillator is just misdirection.

Case 1:04-cv-01371-JJF

Document 529

Filed 09/10/2007

Page 5 of 8

FISH & RICHARDSON P.C.

The Honorable Joseph J. Farnan, Jr. August 31, 2007

Page 5
expert Dr. Horowitz alone has identified no fewer than 30 invalidity contentions. These numbers vastly exceed the number of claims Power Integrations originally asserted in this case and which Fairchild, after stridently objecting to the burden of responding, succeeded in having the Court order reduced to the seven claims eventually tried. As with Fairchild's prior effort to raise more than 80 prior art references, Power Integrations does not have meaningful notice of Fairchild's intended proofs, without which Power Integrations cannot structure its answering case. When asked during his recent deposition whether he intended to opine on all of the 30+ opinions set out in his supplemental report at trial, Dr. Horowitz demurred, asserting that he had no knowledge one way or the other as to what specific contentions he planned to testify to as part of his direct testimony. (Ex. A (Horowitz Tr. 1/30/07) at 15:8-25.) However, with four days allotted for trial, it is clear that Fairchild will not introduce testimony on anywhere near the number of contentions it has enumerated to date, and Power Integrations therefore seeks an order compelling Fairchild to identify the specific contentions it intends to address at trial sufficiently in advance so as to enable Power Integrations to prepare its response.

IV. V.

[WITHDRAWN] THE COURT SHOULD PREVENT FAIRCHILD'S BACKUP TECHNICAL EXPERT, GU-YEON WEI, FROM PROVIDING ANY TESTIMONY IN THIS CASE

Fairchild has included its purported "expert on copying," Gu-Yeon Wei, on its most recent revision of its witness list, but Dr. Wei cannot provide any testimony on "copying" that will assist the finder of fact in this matter. Although the Court denied a similar motion in limine before the first trial, the testimony during the first trial showed that Dr. Wei has no first-hand knowledge of any facts relevant to Fairchild's copying of the claimed inventions, as he freely admitted he has no knowledge of and did not consider Fairchild's process of designing the infringing products in forming his "opinion" on copying. (See Ex. B (Trial Tr.) at 1165-67.) As such, Dr. Wei's opinion cannot meet the standards for admissibility of expert testimony. Further, in concluding that Fairchild willfully infringed, the jury in the first trial implicitly found that Fairchild copied Power Integrations' inventions. Thus, allowing Fairchild to put on evidence to the contrary, especially evidence not directly probative of the question, would undermine the first jury's verdict. In addition, the prejudice to Power Integrations in allowing the Wei testimony is different and greater in the upcoming trial than in the previous proceeding on infringement. This is true because Dr. Wei's "opinion" on copying includes statements about the alleged "similarities" between Power Integrations' patented inventions and the prior art-clearly an improper back-door attempt to bolster Fairchild's invalidity case. But Dr. Wei never provided an expert report on validity, and he cannot, therefore, address this issue at trial. Dr. Wei's proposed testimony on

Case 1:04-cv-01371-JJF

Document 529

Filed 09/10/2007

Page 6 of 8

FISH & RICHARDSON P.C.

The Honorable Joseph J . Farnan, Jr. August 31, 2007

Page 6
the prior art, and his completely baseless assertion that one "could argue" that Power Integrations "copied" the prior art, are thus very likely to confuse the jury, who will interpret his comments as directed to validity. Therefore, pursuant to F.R.E. 402 and 403, the Court should prevent Fairchild from wasting time at trial and potentially confusing the jury with Dr. Wei's testimony. VI. THE COURT SHOULD PRECLUDE FAIRCHILD'S EXPERT DR. PAUL HOROWITZ FROM RELYING ON ALLEGED "PRIOR ART" NOT DISCLOSED IN HIS EXPERT REPORT AND WHICH LACKS FOUNDATION, INCLUDING SPECIFICALLY "DEMONSTRATION MATERIALS" FAIRCHILD MANUFACTURED DURING LITIGATION PURPORTING TO EXHIBIT A MODE OF OPERATION OF AN INTEGRATED CIRCUIT NOT KNOWN BEFORE THIS CASE

In his Supplemental Report, Dr. Horowitz relied on the TEA2262 device as allegedly invalidating prior art to Power Integrations' '851 patent based on the description of a "burst mode" feature in the TEA2262 documentation. However, Dr. Horowitz admitted during his deposition that tests performed after submitting his report confirmed that the documented "burst mode" of the TEA2262 does not, in fact, anticipate any claim of the '851 patent' Rp" Dr. Horowitz opined in his Supplemental Report). (Ex. A at 123:1-16; 129:11'Because the documented burst mode was the sole basis for his invalidity arguments in his Supplemental Report, Dr. Horowitz set forth a new theory of invalidity based on the TEA2262 during his deposition - one not found anywhere in his supplemental expert report. (Id. at 93:14-19; 99:5-9.) This new theory is that the TEA2262 device allegedly invalidates the '851 patent based on a purported "discontinuous burst mode" of operation -- terminology Dr. Horowitz admitted he "made up," for a mode of operation Dr. Horowitz admitted is not described in the TEA2262 documentation.6 To support this new theory, Dr. Horowitz had Fairchild engineers create a "test board" that he now claims, under certain conditions he created in his laboratory tests, allegedly demonstrates "discontinuous burst mode" operation in the TEA2262 device. (Id. at 104:14-113:9.) Dr. Horowitz further admitted during deposition that he has no evidence to suggest that anyone ever operated a TEA2262 device in such a manner before his experiments on the Fairchild-made "test board."
6

REDACTED

Case 1:04-cv-01371-JJF

Document 529

Filed 09/10/2007

Page 7 of 8

FISH & RICHARDSON P.C.

The Honorable Joseph J. Farnan, Jr. August 31, 2007 Page 7

*FnAtTgn
To invalidate a patent, testimony of prior public use or prior invention must generally document the alleged anticipating use or invention, and a prior inventor must have prepared the invalidating invention and known that it worked. Union Carbide Chems. & Plastics Tech. Corp. v. Shell Oil Co., 308 F.3d 1167, 1189-90 (Fed. Cir. 2002) (citations omitted). Here, Dr. Horowitz not only fails to corroborate invalidating prior public use of the TEA2262 product itself, he freely admits having no knowledge of the TEA2262 ever having been used in the manner he alleges anticipates the '851 patent, or of anyone ever having previously built any form of power supply that would exhibit the necessary conditions urt should not permit E® Dr. Horowitz to introduce as "prior art" that which admittedly did not previously exist, especially given that Fairchild's own engineers created the "prior art" during this litigation. Dr. Horowitz's introduction of "demonstration materials" should also be precluded due to the substantial issues of foundation that their use will entail. Given the lack of any evidence of "discontinuous burst mode" prior to Dr. Horowitz and Fairchild's creation of that mode after serving his report, there is no foundation for any testimony regarding this allegedly invalidating feature of the TEA2262. As Dr. Horowitz admitted, this theory is not set forth in his Supplemental Report,

REDACTED
Because Fairchild's only invalidity theory is based on "evidence" not disclosed in an expert report, of a mode of operation for which no evidence ever existed before this litigation, the Court should preclude Fairchild from offering evidence at trial from Dr. Horowitz concerning "discontinuous burst mode" or arguing that the TEA2262 invalidates claims 1 or 4 of the '851 patent based on such an alleged mode of operation.

VII.

THE COURT SHOULD PRECLUDE EXPERT TESTIMONY FROM JAMES BEASOM

In connection with its pretrial submissions, Fairchild has suggested that it intends to have fact witness James Beasom provide previously-undisclosed expert testimony at the upcoming trial. Specifically, Fairchild has stated in its witness list (Pretrial Order at Ex. 9) that Mr. Beasom will testify "that the '173 patent anticipates the '075 patent." Fairchild also produced new prior art earlier this year (DX606-DX613), well after the first trial was completed, and has indicated in its exhibit list (Pretrial Order at Ex. 7) that it intends to use these documents in connection with Mr. Beasom's testimony. Because Fairchild did not designate Mr. Beasom as an expert or provide

Case 1:04-cv-01371-JJF

Document 529

Filed 09/10/2007

Page 8 of 8

FISH & RICHARDSON P.C.

The Honorable Joseph J. Farnan, Jr. August 31, 2007 Page 8 any expert reports from Mr. Beasom, any effort by Fairchild to elicit such testimony from Mr. Beasom would be improper. Respectfully, /s/ William J. Marsden, Jr. William J. Marsden, Jr. Enclosures

WJM/sb
cc: Clerk of Court Steven J. Balick, Esq. (By Hand) Bas de Blank , Esq. (By Federal Express)