`
`__________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`_______________
`
`SONY CORPORATION, SAMSUNG ELECTRONICS CO., LTD.,
`SAMSUNG DISPLAY CO., LTD.
`
`Petitioners,
`
`v.
`
`SURPASS TECH INNOVATION LLC
`
`Patent Owner.
`
`_______________
`
`Case IPR2015-00863
`Patent No. 7,202,843 B2
`
`_______________
`
`SUPPLEMENTAL DECLARATION OF THOMAS L. CREDELLE
`
`
`
`
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`I, Thomas L. Credelle do hereby declare that:
`
`1. I have been retained by counsel for Petitioner Sony Corporation
`
`(“Sony”) to provide assistance regarding US Patent No. 7,202,843 (“the ’843
`
`Patent”).
`
`2. I previously submitted a Declaration in IPR2015-00863 (Ex. 1014)
`
`setting forth my Background and credentials (Credelle Decl. (Ex. 1024) at ¶¶2-14),
`
`and my curriculum vitae (Ex. 1016) which provides further details.
`
`3. I submit this supplemental declaration in response to the Declaration
`
`and Testimony of William K. Bohannon. I incorporate by reference my testimony
`
`set forth my prior Declaration (Ex. 1014).
`
`4. In preparation of this document, I have reviewed the following
`
`documents in addition to those listed in my prior Declaration:
`
` Declaration of William K. Bohannon in Response to Petition of Sony
`
`Corporation et al., identified as Ex. 2022;
`
` U.S. Patent No. 5,642,133 (“Scheffer”) (Ex. 2019);
`
` U.S. Patent No. 5,280,280 (“Hotto”) (Ex. 2020); and
`
` Transcript of January 27, 2016 Deposition of William K. Bohannon,
`
`identified as Ex. 1019.
`
`
`
` P a g e 1 | 17
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`5. In addition, in preparing this document, I have drawn on my
`
`experience and knowledge, as discussed above and described more fully in my CV,
`
`in the areas of LCD technology and flat panel displays.
`
`6. I agree with Mr. Bohannon’s assessment that a person of ordinary
`
`skill in the art in 2003 would have “the ability to understand the overdriving
`
`concept as it is discussed in the ’843 patent.” (See Bohannon Decl. (Ex. 2022) at
`
`¶8.) I would expect this background to include experience in LCD control
`
`electronics. A person having this background would understand factors associated
`
`with driving electronic impulses, and would also understand the concepts of pixel
`
`voltage versus light transmission and pixel response time.” (Id.) I disagree with
`
`Mr. Bohannon that a degree in Mathematics or Computer Science would be an
`
`appropriate educational background in relation to these concepts. (See Bohannon
`
`Decl. (Ex. 2022) at ¶8.)
`
`I.
`
`
`
`Response to Mr. Bohannon’s Comments on My Deposition Testimony
`That the LCD Panel in Suzuki Is an AMLCD
`
`7. In paragraphs 34-40 of Mr. Bohannon’s Declaration, he expresses
`
`disagreement with the testimony that I gave during my deposition on October 28,
`
`2015, in which I described four indicators in Suzuki that Suzuki’s LCD panel is an
`
`active matrix LCD: (1) Suzuki’s use of the terms “source” and “gate” drivers; (2)
`
`Suzuki’s achievement of blur-free moving images at a frame rate of about 60
`
`frames per second, with each frame period lasting 16.6 ms; (3) Suzuki’s use of the
` P a g e 2 | 17
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`term “hold” drive to describe one of its embodiments; and (4) Suzuki’s use of the
`
`term “overdriving.” I will address my Bohannon’s comments on my testimony in
`
`that order.
`
`A. “Source” and “Gate” Drivers
`
`8. The first indicator that Suzuki's liquid crystal panel is an active
`
`matrix liquid crystal display panel and not a passive matrix liquid crystal display
`
`panel is Suzuki’s use of the terms “source” driver and “gate” driver for the data
`
`driver and scan driver respectively. (See Suzuki (Ex. 1003), Fig. 1 reproduced
`
`below.)
`
`
`
`
`
`Suzuki Fig. 1
`
` P a g e 3 | 17
`
`
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`Mr. Bohannon testified that he does not agree that this disclosure indicates “with
`
`certainty” that Suzuki discloses AMLCD. (Bohannon Decl. (Ex. 2022) at ¶ 40.)
`
`However, a person of ordinary skill in the art would have understood that the terms
`
`“source” driver and “gate” driver are applicable to AMLCD panels, and not to
`
`passive matrix panels, and, on that basis, would have assumed that Suzuki’s LCD
`
`panel is an AMLCD.
`
`9. As an initial matter, I agree with Mr. Bohannon that the source driver
`
`16 generates the applied voltage VS to be supplied to the pixels (shown in Suzuki
`
`Fig. 2) and the gate driver 18 generates gate signals GT for selecting pixels.
`
`(Bohannon Decl. (Ex. 2022) at ¶40.) Further Suzuki discloses “a plurality of
`
`pixels P which are formed in a matrix.” (Suzuki (Ex. 1003) at ¶ 47.)
`
`10. An active matrix LCD panel comprises an array of pixels where each
`
`pixel includes a switch such as a TFT; the TFT has a “gate” which is used to open
`
`and close the switch, as well as a “source” and “drain” to allow current to flow to
`
`the liquid crystal capacitor, and optionally a storage capacitor. Passive matrix
`
`LCD panels do not include transistors (nor did they in the 2003 time frame).
`
`
`
` P a g e 4 | 17
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`
`
`Ex. 1010 at p.34
`
`
`
`Ex. 1009 at p. 25
`
`11. Typically, the source of the TFT is connected to a source line which
`
`is connected to a source driver and the drain is connected to the liquid crystal
`
`capacitor (and sometimes an additional storage capacitor), although the terms
`
` P a g e 5 | 17
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`“source” and “drain” are sometimes interchanged due to the symmetry of the TFT
`
`and the type of silicon used. Fig 1-19 from O’Mara (Ex. 1010 at p. 34) reproduced
`
`above illustrates a typical active matrix LCD showing the TFT switches at each
`
`pixel, along with the “gate scan” driver and the “data input” driver. The gate scan
`
`driver is connected to the gate of the TFT and the data input driver is connected to
`
`the source of the TFT. In Fig. 2.16 from Lueder (Ex. 1009 at p. 25) reproduced
`
`above, the TFTs for two pixels are shown. A gate impulse Vg from a “gate” driver
`
`(not shown) is applied to the n-th row and Vvideo from a “source” or data driver (not
`
`shown) is applied to the liquid crystal capacitor and storage capacitor through the
`
`source/drain of the TFTs. I agree with Mr. Bohannon’s testimony stating in
`
`reference to drivers for active matrix LCDs: “We’re driving an active matrix, we’re
`
`going to use a gate driver, it’s a low voltage driver, it works like this.” (Bohannon
`
`Tr. (Ex. 1019) at 144:2-4 (emphasis added).) Therefore, a person of ordinary skill
`
`in the art (at the time of the alleged invention) would have assumed that the circuit
`
`and driving waveforms of Suzuki Fig 1 were intended to be applied to an active
`
`matrix LCD despite the lack of a detailed description of active matrix elements in
`
`Suzuki.
`
`B. “Moving Images”
`
`12. The second indicator of applicability of Suzuki to active matrix
`
`LCDs is its stated goal to “improve the moving image display performance of a
`
` P a g e 6 | 17
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`liquid crystal display device.” (Suzuki (Ex. At 1003) ¶ 8.) Mr. Bohannon testified
`
`that he agrees with my testimony “that passive LCD technology was capable of
`
`displaying moving images at the time Suzuki was filed.” (Bohannon Decl. (Ex.
`
`2022) at ¶ 35). To the extent that Mr. Bohannon is characterizing my testimony as
`
`suggesting that passive LCD technology was capable of displaying “moving
`
`images” within the meaning of that term as Suzuki uses it, he is incorrect and has
`
`taken my testimony out of context. When asked whether “passive matrix LCD
`
`panels have ever been used for television,” I responded: “Not to my knowledge.”
`
`Credelle Tr. (Ex. 2004) at 27:16-18 (emphasis added). And when asked further
`
`whether passive matrix LCDs have “ever been used for computer monitors that are
`
`required to display moving images, I responded: “They are used for computer
`
`monitors, and the computer can’t dictate what content the user may try to display.
`
`But the response time of a passive matrix LCD is too slow for motion video.” Id.
`
`at 27:21 – 28:1 (emphasis added). Suzuki is concerned with “moving images”
`
`displayed by LCD panels used for “personal computers, television sets, and so on.”
`
`Suzuki (Ex. 1001) at ¶ 0004 (emphasis added). Consistent with the object of
`
`displaying full motion video (e.g., television), Suzuki’s preferred embodiment
`
`drives an LCD that receives video images at 60 frames per second. (See Suzuki
`
`(Ex. 1003) at ¶ 39.)
`
`
`
` P a g e 7 | 17
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`13. A person of ordinary skill in the art would have assumed from
`
`Suzuki’s frame rate, which equates to a frame period of 16.6 ms (Suzuki (Ex.
`
`1003) at ¶ 39), that an active matrix LCD with fast response liquid crystal material
`
`is used (e.g., less than or equal to 16.6 ms). In 2003 time frame, no 640x480 gray-
`
`scale-capable passive matrix displays could achieve this goal. In fact, passive
`
`matrix displays such as STN-LCDs with line-at-a-time addressing required a much
`
`longer response time, e.g., 100s of ms, in order to prevent display flicker. (See
`
`e.g., Lueder (Ex. 1009) at pp. 176-177.) To solve the response time issue in
`
`passive matrix displays, and allow the use of faster LC materials, “multi-line-
`
`addressing” was developed. (See, e.g., Ex. 2019 and Ex. 2020.) The complex
`
`multi-line-addressing waveforms are quite different from the simple addressing
`
`waveform of Suzuki Fig. 2, thus a person of ordinary skill in the art would not
`
`assume that Suzuki is referring to any multi-line-addressing circuit solution. Even
`
`with these new addressing schemes, the response time of the passive matrix LCDs
`
`were on the order of 50-60 ms (Ex. 2019 at 3:50-56 and Ex. 2020 at 4:34-35),
`
`which is still much longer than Suzuki’s frame time of 16.6 ms. While Mr.
`
`Bohannon testified that some small 240 line Casio TVs that utilized passive matrix
`
`LCDs “produced excellent video images” (Bohannon Tr. At 164:5-10.), if such
`
`TVs existed, they were uncommon, and certainly not the type of television that
`
`
`
` P a g e 8 | 17
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`would have come to mind to a person of ordinary skill in the art when considering
`
`Suzuki’s LCDs for PC and television use.
`
`14. Given Suzuki’s frame rate, and that Suzuki’s invention is directed to
`
`LCD’s for “personal computers, televisions, and so on,” a person of ordinary skill
`
`in the art would have assumed that Suzuki’s LCD panel is an active matrix panel,
`
`and not a passive matrix panel. (See Suzuki (Ex. 1003) at ¶¶ 4, 39.).
`
`C. “Hold drive”
`
`15. The third indicator of the applicability of Suzuki to active matrix
`
`LCD drives is Suzuki’s use of the term “hold drive” in describing one of its
`
`embodiments. In his declaration, Mr. Bohannon testified that the term “hold drive”
`
`simply refers to the use of frame buffers with LCDs. (Bohannon Decl. (Ex. 2022)
`
`at ¶38). I disagree. Suzuki refers to a “hold drive” to distinguish the LCD panels
`
`to which its technology is applied from CRTs. (Suzuki (Ex. 1003) at ¶ 5). As Mr.
`
`Bohannon testified, frame buffers are used with CRTs in some cases, and
`
`therefore, in referring to a “hold drive,” Suzuki cannot mean a frame buffer
`
`because the use of a frame buffer is not a point of distinction between an LCD
`
`panel and a CRT. (See Bohannon Tr. (Ex. 1019) at 126:7-16.) A person of
`
`ordinary skill in the art would not have understood that the term “hold drive” in
`
`Suzuki to mean a frame buffer for that reason.
`
`
`
` P a g e 9 | 17
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`16. Instead, a person of ordinary skill in the art would have assumed that
`
`a “hold drive” refers to an AMLCD panel. The context in which the term “hold
`
`drive” is used in the description of Suzuki’s preferred embodiment supports the
`
`understanding that an AMLCD is meant. Suzuki states that “the liquid crystal
`
`display device of this embodiment operates on hold drive. That is, data signals
`
`corresponding to the same image data are supplied to the liquid crystal cells over a
`
`period of one frame (16.6 ms).” (Suzuki (Ex. 1003) at ¶ 39). In order for the data
`
`signals, which are the analog drive voltages, to be held for a frame time, there must
`
`be a switch at the pixel, i.e., a thin film transistor (“TFT”) of an active matrix,
`
`which allows the data voltage to be established on the LC and storage capacitor
`
`when the switch is closed and then held on the capacitor when the switch is open.
`
`That is, when a gate signal is applied to a row, the TFT switch closes and the
`
`source driver supplies current to charge up the liquid crystal capacitor to the
`
`desired data voltage. When the gate signal is turned off, the charge remains on the
`
`liquid crystal until addressed again. Thus, the voltage is “held” on the pixel.
`
`Suzuki’s Fig. 2 (reproduced above) shows the applied voltage VS is held on the
`
`pixel at level (a) for the first half frame SF1 and level (c) for the second half frame
`
`SF2. (Suzuki (Ex. 1003) at Fig. 2.) During his deposition, Mr. Bohannon agreed
`
`that Suzuki shows that voltages are held on the pixel for a large fraction of the
`
`frame time (minus any losses due to charge leakage). (See Bohannon Tr. (Ex.
`
` P a g e 10 | 17
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`1019) at 31:22-33:3.) Mr. Bohannon also testified that “active matrix has got a
`
`duty cycle of 1.” (Bohannon Tr. (Ex. 1019) at 47:4-21.) A duty cycle of 1 means
`
`that the voltage is held on the pixel for an entire frame.
`
`17. In contrast, a passive matrix LCD has no means to “hold” the voltage
`
`on a pixel; the driving voltage is applied to a pixel as an “impulse” for one line
`
`time, e.g., 1/240 of a frame period (approximately 69 µs for a frame time of
`
`16.6ms) by applying a high voltage pulse to the row electrode and a data voltage to
`
`the column electrode. The liquid crystal molecules respond to the rms (root mean
`
`square) voltage of the voltages applied over a frame period, as illustrated in Figs. 1
`
`and 3B of Scheffer (reproduced below). (See Scheffer (Ex. 2019) at Figs. 1 and
`
`3B.) Fig. 1 shows the applied row and column voltages for the first 5 rows and
`
`two columns. (See Scheffer (Ex. 2019) at Fig. 1.). For example, pixel 265 is
`
`addressed with row voltage S and column voltage +D and results in a dark pixel;
`
`pixel 261 is addressed with row voltage S and column voltage -D and results in a
`
`bright pixel. (Id.) The voltage applied over the first 5 line times is shown in Fig.
`
`3B; it is the rms voltage of this waveform over a complete frame that drives the
`
`liquid crystal to a dark or light state. (See Scheffer (Ex. 2019) at Fig. 3B.) Since
`
`there is no hold voltage, the liquid crystal material is chosen to have a response
`
`time longer than the frame time in order to prevent flicker. In fact, I agree with
`
`Mr. Bohannon’s testimony that passive matrix LCDs are “energized” row by row.
`
` P a g e 11 | 17
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`(Bohannon Tr. (Ex. 1019) at 43:19-44:21.) But I disagree with Mr. Bohannon’s
`
`testimony that there is any voltage held on the LC capacitance in a passive matrix
`
`LCD panel. (See Bohannon Tr. (Ex. 1019) at 50:11-51:19). In a passive matrix
`
`LCD panel, the voltage on a pixel is constantly changing. In such a panel, the
`
`LCD responds to the rms voltage applied to the pixel averaged over the entire
`
`frame time and there is no voltage “held” on the LC.
`
`
`
`Scheffer Fig. 1
`
`
`
` P a g e 12 | 17
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`
`
`Scheffer Fig. 3B
`18. From this analysis, a person of ordinary skill in the art would
`
`conclude that the drive circuit of Suzuki applies to an active matrix, and not a
`
`passive matrix LCD. Further, a person of ordinary skill in the art would not know
`
`how to apply the waveforms of Suzuki to a passive matrix LCD as there are no
`
`means to hold the voltage for a large fraction of a frame period.
`
`D. “Overdriving”
`
`19. At the time of both the ’843 Patent and the Suzuki disclosure, I am
`
`not aware of any “overdrive” solutions for passive matrix LCDs, which is defined
`
`in the ’843 Patent as “applying a higher or a lower data impulse to the pixel
`
`electrode to accelerate the reaction speed of the liquid crystal molecules, so that
`
`the pixel can reach the predetermined gray level in a predetermined frame
`
`period.” (’843 Patent (Ex. 1001) at 2:1-7 (emphasis added).) In Mr. Bohannon’s
`
`declaration (Ex. Sony-2022 at ¶34), he cited two patents, Scheffer (Ex. 2019) and
`
` P a g e 13 | 17
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`Hotto (Ex. 2020) that describe driving methods for passive matrix LCDs as
`
`evidence that overdrive had been applied to passive matrix LCDs. (Bohannon
`
`Decl. (Ex. 2022) at ¶ 34). I disagree with Mr. Bohannon that these patents disclose
`
`any form of overdrive as defined in the ’843 Patent. These patents describe
`
`alternate methods of controlling the transmission rate, or transmittance, to achieve
`
`more accurate gray scale levels on passive matrix LCDs (Scheffer) and both active
`
`and passive matrix LCDs (Hotto), but do not describe overdriving. (See Scheffer
`
`(Ex. 2019) and Hotto (Ex. 2020).)
`
`20. Scheffer describes a new driving scheme to achieve more precise
`
`gray scales in an STN LCD at lower drive frequencies and without using pulse
`
`width modulation. (Scheffer (Ex. 2019).) In Scheffer, this is achieved by splitting
`
`the line address time (typically 60-70 µs for a dual scan 640x480 display) in half
`
`and using two different analog drive voltages. (Scheffer (Ex. 2019) at 7:50-55 and
`
`Fig. 3B.) This drive scheme has no impact on acceleration of the reaction speed of
`
`the liquid crystal molecules.
`
`21. Mr. Bohannon has testified that Suzuki’s Fig. 5 is comparable to
`
`Scheffer’s Fig. 3B. (See Bohannon Decl. (Ex. 2022) at 42.) I disagree. Suzuki’s
`
`Fig. 5 illustrates the voltage applied to a pixel where the voltage is constant for ¼
`
`and ¾ of a frame, which is equivalent to approximately 4.15 and 12.45 ms for a
`
`60Hz refresh display (typical of computer monitors and TVs in the early 2000 time
`
` P a g e 14 | 17
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`frame). During his deposition, Mr. Bohannon testified that the overdrive pulses
`
`SF1 and SF2 of Suzuki Fig. 5 (4.15 ms and 12.45 ms respectively) are somehow
`
`equivalent to the pulse widths Δs1 and Δs2 of Scheffer’s Fig. 3B. These Δs pulses
`
`are equivalent to 0.5*(frame time/240) (for a dual scan 640x480 passive matrix
`
`display), or approximately 35 µs for a 16.6ms frame period—this is more than 100
`
`times smaller than SF1 or SF2. (See Bohannon Tr. (Ex. 1019) at 68:9-69:5.)
`
`Since Scheffer’s disclosure is aimed at matrix displays having 480 rows, that
`
`Suzuki’s time intervals may be more similar to Scheffer’s when considering
`
`“something with fewer pixels,” is irrelevant. (See Bohannon Tr. (Ex. 1019) at
`
`68:16-69:5.)
`
`22. The second patent upon which Mr. Bohannon relies as allegedly
`
`showing overdriving in passive matrix displays is Hotto (Ex. 2020). (Bohannon
`
`Decl. (Ex. 2022) at ¶ 34). Hotto discloses a complex addressing scheme applicable
`
`to both active matrix and passive matrix LCDs that uses real time computation and
`
`memory circuits to apply asynchronous voltage pulses to the LCD in better achieve
`
`gray levels and avoid DC bias effects. (Hotto (Ex. 2020).) Hotto uses the term
`
`“Pixel Power Modulation,” which Mr. Bohannon characterizes as “overdrive”.
`
`(Bohannon Decl. (Ex. 2022 at ¶ 34)). However, Hotto does not define Pixel Power
`
`Modulation to mean “overdrive.” Instead, Hotto defines it as “a control
`
`technique…in which the display controller selective varies (or modulates) in real
`
` P a g e 15 | 17
`
`
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`time the power applied to individual pixels to maintain them in the desired gray
`
`band.” (Hotto (Ex. 2020) at 2:40-44). Mr. Bohannon also cites Fig. 6 of Hotto as
`
`evidence of “overdrive.” (Bohannon Decl. (Ex. 2022) at ¶ 42.) However, Fig. 6 of
`
`Hotto merely shows how “many drive pulses applied in rapid succession” can
`
`achieve a more precise gray level. (Hotto (Ex. 2020) at 11:37-38.) Thus, Hotto’s
`
`drive technique is another way to achieve better gray scale control, and nowhere in
`
`the patent does it describe any methods or circuits to accelerate of the reaction
`
`speed of the liquid crystal molecules, as required by “overdrive” as the term in
`
`used in the ’843 Patent.
`
`23. Thus, both Scheffer and Hotto are alternate drive techniques to
`
`achieve better control of the transmittance (or transmission rate) to achieve better
`
`gray scale control of an LCD. I do not agree with Mr. Bohannon’s testimony that
`
`these are forms of “overdrive”.
`
`24. Taking all of the four indicators detailed above, a person of ordinary
`
`skill in the art would assume that the LCD described by Suzuki is an active matrix
`
`LCD, even though a detailed drawing is not shown. It would have been obvious to
`
`combine Suzuki and Nitta, since both address the same blurring issue with
`
`overdrive circuitry and Nitta clearly describes an active matrix LCD.
`
`
`
`
`
`
`
` P a g e 16 | 17
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020
`
`
`
`I declare under penalty of perjury under the laws of the United States of
`
`America that the foregoing is true and correct.
`
`Dated:
`
`9»
`
`/9
`
`
`
`Thoma; L. Credelle
`
`Page 17|17
`
`SONY V. SURPASS Tech., |PR2015-00863
`Exhibit SONY-1020
`
`
`
`SONY v. SURPASS Tech., IPR2015-00863
`Exhibit SONY-1020