THE UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`DAIHEN CORPORATION,
`Petitioner,
`
`v.
`
`RENO TECHNOLOGIES, INC.,
`Patent Owner.
`
`Case IPR2019-00248
`Patent 9,496,122
`
`PATENT OWNER PRELIMINARY RESPONSE PURSUANT TO 35 U.S.C.
`§ 316 AND C.F.R. 37 § 42.107 TO PETITION FOR INTER PARTES
`REVIEW OF U.S. PATENT NO. 9,496,122
`
`Page 1 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`LIST OF PATENT-OWNER EXHIBITS
`
`Exhibit No.
`
`Description
`
`2001
`
`2002
`
`2003
`
`2004
`
`2005
`
`2006
`
`2007
`
`2008
`
`Mark Andrews, Reno Sub-Systems Sets Pace in Plasma
`Process Control, Silicon Semiconductor (Vol. 39 Issue 4
`2017) (“Silicon Semiconductor”).
`Dylan McGrath, Samsung, Intel Back Process Control
`Vendor, EE Times (Sep. 28, 2017),
`https://www.eetimes.com/document.asp?doc_id=1332371
`(“EE Times”).
`U.S. Patent No. 5,654,679 (“Mavretic”).
`
`U.S. Patent No. 6,887,339 (“Goodman”).
`
`U.S. Patent No. 7,030,717 (“Chung”) .
`
`Paramount Series (Data Sheet), Advanced Energy
`Industries, Inc. (2016),
`https://www.advancedenergy.com/globalassets/resources-
`root/data-sheets/paramount-series-data-sheet.pdf, (“AE
`Generator Data Sheet”).
`GHW12Z (Specification), MKS Instruments, Inc.,
`https://www.johnmorrisgroup.com/Content/
`Attachments/124386/GHW12Z-specifications.pdf (“MKS
`Generator Specification”).
`GHW50A RF Generator (Specification), ENI (Division of
`Astec America, Inc.),
`https://www.johnmorrisgroup.com/Content/
`Attachments/124386/GHW50A-specifications.pdf (“ENI
`Generator Specification”).
`
`
`
`
`
`
`
`
`
`Page 2 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`I.
`II.
`
`V.
`
`TABLE OF CONTENTS
`Introduction ........................................................................................... 5
`The ‘122 Patent ..................................................................................... 8
`Background: Semiconductor fabrication and the role of the
`matching network and the variable capacitor ............................. 8
`Electronically variable capacitors and inventor Imran
`Bhutta ........................................................................................ 10
`Determining the variable impedance of the plasma chamber ... 14
`Timing Feature .......................................................................... 15
`Capacitor tuning versus prior-art frequency tuning and other
`types of RF generator signal tuning .......................................... 16
`III. Claim construction .............................................................................. 18
`IV. The Petition’s grounds of supposed rejection identified in the
`Petition’s table are different from the grounds described in the
`body of the Petition and this response uses the nomenclature of
`the body rather than the table. ............................................................. 19
`Summary of Zhang and Chen .............................................................. 19
`Zhang (Ex. 1006) ...................................................................... 20
`1.
`Substrate process system ................................................ 20
`2.
`Error feedback-based tuning methods ............................ 22
`3.
`Fast frequency tuning ..................................................... 25
`Chen (Ex. 1008) ........................................................................ 26
`VI. Relevant Law ....................................................................................... 29
`Standard of Review ................................................................... 29
`Anticipation and obviousness ................................................... 30
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`- 2 -
`
`Page 3 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`
`
`
`
`
`
`
`
`VII. The Board should reject all Zhang-based grounds – namely,
`grounds 1, 2 and 5 – because the Petition fails to establish a
`likelihood of success as to any of the challenged claims. ................... 31
`The Board should deny institution as to all Zhang-based
`grounds because Patent Owner cited Zhang during original
`prosecution and the Petition nowhere provides explanation
`showing why original allowance was incorrect. ....................... 32
`Zhang does not disclose or suggest “determining” variable
`impedance of a plasma chamber but instead uses error
`feedback involving incrementally adjusting impedance until
`reflected power is reduced or eliminated. ................................. 33
`Zhang does not disclose or suggest the Timing Feature of
`the challenged claims because Zhang never determines the
`chamber impedance in the first place and, in any event,
`nowhere teaches or suggests using variable capacitors to
`meet the Timing Feature. .......................................................... 40
`The Petition’s allegations that Zhang renders challenged
`claims obvious based on the knowledge of the ordinarily-
`skilled artisan or on Chen should be rejected because the
`Petition relies on Zhang for the “determining” limitations
`and the Timing Feature discussed above. ................................. 42
`VIII. The Board should reject all Chen-based grounds – namely,
`grounds 3-6 – because the Petition fails to establish a likelihood of
`success as to any of the challenged claims. ......................................... 43
`The Board should reject ground 5, Zhang combined with
`Chen, because the Petition relies on Zhang (and not Chen)
`as disclosing determining variable chamber impedance, and
`for the reasons provided above, Zhang does not disclose or
`suggest such determining. ......................................................... 43
`The Board should reject grounds 3, 4, and 6 involving
`alleged obviousness combinations based on Chen because
`Chen does not teach or suggest the Timing Feature. ................ 44
`The Board should reject grounds 3, 4, and 6 involving
`alleged obviousness combinations based on Chen or
`- 3 -
`
`
`
`
`
`
`
`
`
`
`Page 4 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`Howald because neither discloses use of series-based
`electronically variable capacitors. ............................................. 48
`IX. Conclusion ........................................................................................... 49
`
`
`
`- 4 -
`
`
`
`
`
`
`
`Page 5 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`I.
`
`Introduction
`Petitioner is wrong—the invention of the ‘122 patent stands neither
`
`anticipated nor obvious. Rather, the invention represents an important innovation
`
`in the technology of impedance matching networks.
`
`Patent Owner Reno is a small start-up company having about 30 employees.
`
`Its new, novel, nonobvious and commercially successful impedance-matching
`
`technology is shaking up the semiconductor manufacturing industry, leaving
`
`competitors like Petitioner scrambling and losing business. But instead of
`
`competing in the marketplace, Petitioner – a large multinational corporation – now
`
`attempts an end-run by thrusting irrelevant prior art before the Board and hoping
`
`the complicated
`
`technology offers camouflage for
`
`ill-conceived
`
`invalidity
`
`allegations. Patent Owner respectfully requests that the Board nip this attempt in
`
`the bud and see the Petition for what it really is – a poorly thrown Hail Mary.
`
`Impedance matching networks maximize power transfer between an RF
`
`source and a plasma chamber. The matching network of the ‘122 patent achieves
`
`unprecedented matching speed using a capacitor-based matching network as
`
`opposed to matching networks using, whether alone or with capacitor matching,
`
`frequency-tuning methods. The invention uses electronically variable capacitors
`
`(EVCs) in both shunt and series, enabling unprecedented matching speed in a
`
`capacitor-based matching network. Further, by using EVC-based capacitor tuning
`
`- 5 -
`
`Page 6 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`and not frequency tuning, the ‘122 patent’s network provides matching over a wide
`
`range of impedances, unlike the narrower range available with frequency tuning.
`
`Because of its impedance-range limitation, fast frequency-tuning matching alone is
`
`typically insufficient for providing impedance matching, hence why it is often
`
`paired with some sort of capacitor-based matching. But the invention embodied in
`
`the challenged claims overcomes the impedance-range limitation of frequency-
`
`tuning matching systems by using capacitors for matching while achieving
`
`unprecedented frequency-tuning-like speeds using its shunt-and-series EVC
`
`arrangement.
`
`All proposed invalidity grounds in the Petition rely on either Zhang or Chen.
`
`But Zhang was cited to the PTO during original prosecution, and the examiner
`
`allowed the claims over Zhang likely for one or both of two reasons. First, Zhang
`
`does not determine a variable impedance of a plasma chamber in the matching
`
`process, as recited in all challenged claims. Instead, Zhang uses an incremental
`
`trial-and-error approach involving cycles of: checking for reflected power,
`
`adjusting impedance, checking for reflected power, adjusting impedance, and
`
`continuing this until reflected power is below a predetermined threshold. Unlike
`
`the invention of the challenged claims, Zhang does not determine any impedance.
`
`Second, Zhang does not disclose the Timing Feature of the ‘122 patent—
`
`namely, that in less than about 150 µsec an EVC-based matching network goes
`
`
`
`
`- 6 -
`
`Page 7 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`from determining the plasma-chamber impedance to reducing reflected RF power.
`
`Again, Zhang never determines the plasma chamber impedance in the first place.
`
`And while Zhang does reference “100 microseconds,” the Petition fails to explain
`
`that this is referring not to impedance matching using capacitor tuning, but to
`
`impedance matching using well-known frequency tuning – the only method for
`
`impedance matching in the prior art to achieve such speeds. But unlike the
`
`invention’s shunt-and-series EVC configuration that offers fast matching over a
`
`wide-range of impedances, frequency matching provides matching over only a
`
`narrow range.
`
`Chen is equally deficient for two reasons. First, Chen is primarily directed
`
`to impedance matching using frequency, not capacitor (let alone EVC), tuning.
`
`Thus, it nowhere discloses the recited Timing Feature using capacitor tuning.
`
`Instead, it references 100-µsec timing with respect to only frequency tuning, which
`
`again was a well-known feature of the prior art.
`
`Second, Chen does not disclose using EVCs in both the series and shunt
`
`positions but instead discloses only using mechanically variable capacitors in the
`
`series position. This more readily supports that Chen nowhere comes close to the
`
`speed of the invention of the challenged claims using capacitor-based matching.
`
`
`
`
`
`
`- 7 -
`
`Page 8 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`II. The ‘122 Patent
` Background: Semiconductor fabrication and the role of the
`matching network and the variable capacitor
`Plasma processing is used to fabricate semiconductor devices. Ex. 1001
`
`(‘122 patent), col. 1, ll. 21-65. Plasma processing involves energizing a gas
`
`mixture by imparting energy to the gas molecules by introducing radio frequency
`
`(RF) energy into the gas mixture using an RF source (also referred to as an RF
`
`generator). Id. This gas mixture is typically contained in a vacuum chamber,
`
`referred to as the plasma chamber. Id.
`
`The RF source has a fixed impedance of about 50 Ohms. Id. By contrast,
`
`the impedance of the plasma chamber varies during the semiconductor fabrication
`
`process. Id. The differing impedances of the RF source and the plasma chamber
`
`cause reflected RF power, and thus inefficient transmission of the RF power from
`
`the RF source to the plasma chamber. Id.
`
`An impedance matching network is situated between the RF source and
`
`plasma chamber to transform the impedance of the plasma chamber to 50 Ohms,
`
`thus causing an impedance match between the RF source and the plasma chamber.
`
`Id. A simplified block diagram of the semiconductor fabrication system of the
`
`‘122 patent is provided below.
`
`
`
`
`
`
`
`
`- 8 -
`
`Page 9 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`
`
`
`
`
`RF source
`15
`
`Impedance
`matching
`network
`11
`
`Semiconductor Fabrication System
`
`Plasma
`chamber
`19
`
`See id., Fig. 1.
`
`As the plasma chamber’s impedance changes, the matching network can
`
`change to maintain the impedance match. Id., col. 2, ll. 10-17. The matching
`
`network can include variable capacitors for changing the matching network’s
`
`impedance, and thus causing an impedance match. Id. And as discussed below,
`
`impedance matching can also or alternatively be carried out by altering the RF
`
`signal output by the RF generator.
`
`The predominant type of variable capacitor used in RF impedance matching
`
`networks is the vacuum variable capacitor (VVC). Id., col. 2, ll. 2-46. The VVC is
`
`an electromechanical device having two concentric metallic rings that are moved in
`
`relation to each other to change capacitance. Id. In complex semiconductor
`
`fabrication processes using plasma chambers, where the impedance changes are
`
`frequent, VVCs often experience mechanical failures within less than a year of use.
`
`Id. Such failures lead to downtime for fabrication equipment so that the failed
`
`VVC can be replaced. Id.
`
`As semiconductor devices shrink in size and become more complex, the
`
`feature geometries become smaller, and the processing time for each individual
`- 9 -
`
`
`
`
`Page 10 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`step needed to fabricate these small features is likewise reduced. Id. Matching
`
`networks using VVCs generally take 1-2 seconds to match the plasma chamber
`
`impedance to the RF source impedance. Id. For a significant time during the
`
`matching process, the fabrication process parameters are unstable and must be
`
`accounted for as part of the overall fabrication process. Id.
`
`
`Electronically variable capacitors and inventor Imran Bhutta
`In view of the deficiencies of VVCs, there was need for the development of
`
`a solid-state variable capacitor that did not rely on moving parts to vary
`
`capacitance. Id., col. 2, ll. 2-46. The inventor, Mr. Bhutta, has been a leading
`
`innovator in this field. See, e.g., Ex. 1001; Ex. 1009. For example, his U.S. Patent
`
`No. 7,251,121, developed for his prior company, Innovation Engineering LLC,
`
`describes a novel and innovative capacitive array for varying capacitance. See Ex.
`
`1009 (Bhutta121). This capacitive array is referred to by Reno as an
`
`“Electronically Variable Capacitor” or “EVC”. Ex. 1001 (‘122 patent), col. 2, ll.
`
`38-40.
`
`Rather than using moving parts, each EVC is made of numerous discrete
`
`capacitors connected in parallel. Ex. 1009 (Bhutta121), col. 3, ll. 10-17. The
`
`capacitance of the EVC is adjusted by altering which discrete capacitors are
`
`switched in. Id., col. 5, ll. 22-35. Since there are no moving parts, EVCs can
`
`
`
`
`- 10 -
`
`Page 11 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`switch much faster than mechanically variable capacitors. Ex. 1001 (‘122 patent),
`
`col. 15, Table 1.
`
`Reno Sub-Systems has used Inventor Bhutta’s EVC technology and
`
`subsequent matching network advances
`
`to develop
`
`the revolutionary RF
`
`impedance matching network claimed in the ‘122 patent. See id. While VVC-
`
`based matching networks typically take 1-2 seconds to impedance match, the
`
`matching network of the ‘122 patent can achieve an impedance match in 500 µs or
`
`less. Id., col. 14, l. 24 to col. 15, l. 20. Within 150 µs or less, the matching
`
`network determines the variable impedance of the plasma chamber, determines
`
`new capacitor values, alters the capacitances of the EVCs, waits for capacitor
`
`latency as the added discrete capacitors of the EVCs charge, and causes the
`
`reflected RF power to decrease. Id., col. 13, l. 50 to col. 14, l. 33. By impedance
`
`matching using EVCs instead of VVCs or other mechanically variable capacitors,
`
`the matching network was able to perform its task more quickly and reliably. Id.,
`
`col. 15, ll. 5-65. EVCs also provided several other benefits, including better
`
`current handling and reduced size. Id.
`
`The semiconductor manufacturing
`
`industry quickly
`
`recognized
`
`the
`
`revolutionary qualities of Reno’s EVC-based impedance matching network as
`
`described in the ‘122 patent. See, e.g., Ex. 2001 (Silicon Semiconductor); Ex.
`
`2002 (EE Times). Publications such as Silicon Semiconductor have reported on
`
`
`
`
`- 11 -
`
`Page 12 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`how Reno is “changing plasma processing tools in a big way.” Ex. 2001 (Silicon
`
`Semiconductor) at 24. It reports Reno has managed to “shake up the rather staid
`
`RF power, match and gas flow segments of the plasma world,” capturing investors
`
`such as Intel, Lam Research, and Samsung, and selling their products to “nearly 80
`
`percent of the industry’s largest manufacturers ….” Id. Silicon Semiconductor
`
`goes into detail describing the speed improvements enabled by Reno’s EVC-based
`
`matching network. Id. 25-26. Other publications, such as EE times provide
`
`similar information regarding Reno’s dramatically increased matching speeds and
`
`high-profile investors. Ex. 2002 (EE Times).
`
`Claim 1 of the ‘122 patent is exemplary. It claims a matching network that
`
`includes a series EVC, a shunt EVC, and a control circuit. Ex. 1001 (‘122 patent),
`
`claim 1. Non-limiting Fig. 1 provides an example of such an arrangement, where
`
`series EVC 31 and shunt EVC 33 are arranged in a manner sometimes referred to
`
`as an “L” topology. Id., Fig. 1.
`
`
`
`
`- 12 -
`
`Page 13 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`Fig. 1 of ‘122 Patent
`
`
`
`Claim 1 provides that the control circuit is configured to:
`
`determine the variable impedance of the plasma chamber,
`
`determine a first capacitance value for the first variable capacitance
`and a second capacitance value for
`the second variable
`capacitance, and
`
`generate a control signal to alter at least one of the first variable
`capacitance and the second variable capacitance to the first
`capacitance value and the second capacitance value, respectively,
`
`wherein an elapsed time between determining the variable impedance
`of the plasma chamber to when RF power reflected back to the RF
`source decreases is less than about 150 µsec.
`
`
`Id., claim 1.
`
`
`
`
`- 13 -
`
`Page 14 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
` Determining the variable impedance of the plasma chamber
`As indicated above, claim 1 recites that the RF impedance matching network
`
`comprises a “control circuit configured to: determine the variable impedance of
`
`the plasma chamber.” Id., claim 1 (emphasis added). The variable impedance of
`
`the plasma chamber is sometimes referred to herein as the “chamber impedance.”
`
`Claim 1 further recites that the control circuit determines first and second
`
`capacitance values and alters the series and shunt EVCs to have these respective
`
`capacitance values. Id. The claim concludes, “wherein an elapsed time between
`
`determining the variable impedance of the plasma chamber to when RF power
`
`reflected back to the RF source decreases is less than about 150 μsec.” Id., claim
`
`1. All other challenged claims include similar limitations. Id., cols. 16-18.
`
`In one embodiment of the matching network of the ‘122 patent, “the first
`
`approximately 50 μsec of the match tune process is used by the control circuit to
`
`perform measurements and calculations to determine new values for the variable
`
`capacitances of the series and shunt EVCs.” Id., col. 13, ll. 51-55. These
`
`measurements and calculations include a determination of the chamber impedance,
`
`which can vary over time:
`
`As part of this initial step of the process, the control circuit uses the
`power output of the RF source, the current of the RF signal, and the
`known settings of the series and shunt variable capacitors to
`determine the current variable impedance of the plasma chamber.
`
`
`
`
`- 14 -
`
`Page 15 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`Once the variable impedance of the plasma chamber is known, the
`control circuit can then determine the changes to make to the variable
`capacitances of one or both of the series and shunt variable capacitors
`for purposes of achieving an impedance match.
`
`Id., col. 13, ll. 55-64 (emphasis added). The determination of the chamber
`
`impedance enables the control circuit to determine the changes to make to the
`
`EVCs to directly pursue an impedance match, and thus to achieve the recited rapid
`
`reduction in reflected RF power. Id., col. 13. ll. 55-64. Thus, rather than making
`
`minor adjustments to the variable capacitors and assessing how the changes reduce
`
`reflected RF power to gradually arrive at an impedance match, the claimed method
`
`allows the matching network to immediately adjust the variable capacitors to
`
`capacitance values for providing an impedance match. Id.
`
` Timing Feature
`By enabling the use of an EVC in series, the claimed matching network was
`
`able to achieve high-speed and reliable tuning through the use of not one, but two
`
`EVCs. See Ex. 1001 (‘122 patent), col. 13, l. 26 to col. 14, l. 51; Fig. 9. Regarding
`
`the high-speed nature of the claimed matching network, each of the claims of the
`
`‘122 patent include the following limitation: “wherein an elapsed time between
`
`determining the variable impedance of the plasma chamber to when RF power
`
`reflected back to the RF source decreases is less than about 150 μsec.” Id., cols.
`
`16-18. This limitation is referred to herein as the “Timing Feature.” Achieving the
`
`
`
`
`- 15 -
`
`Page 16 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`Timing Feature by adjusting the values of one or more variable capacitors
`
`(“capacitor tuning”) was, at the time of the invention, unprecedented. Id., col. 14,
`
`ll. 46-51; col. 14, Table 1.
`
`Capacitor tuning versus prior-art frequency tuning and other
`types of RF generator signal tuning
`RF generator signal tuning is a method of impedance matching that is
`
`distinct from the capacitor tuning method described in the ‘122 patent. See Ex.
`
`2003 (Mavretic), col. 2, l. 54 to col. 3, l. 6. While capacitor tuning is carried out
`
`by adjusting the capacitance of one or more variable capacitors to achieve an
`
`impedance match (Id., col. 6, ll. 21-46), RF generator signal tuning is carried out
`
`by adjusting the RF signal being transmitted by the RF generator to achieve an
`
`impedance match (Id., col. 2, l. 54 to col. 3, l. 6).
`
`There are several methods of RF generator signal tuning. For example,
`
`Chen discusses varying the RF frequency or pulse height of the generator signal.
`
`Ex. 1008 (Chen); Id., col. 2, ll. 30-36. The Mavretic patent also discusses
`
`frequency tuning. Ex. 2003 (Mavretic), Abstract.
`
`RF generator signal tuning is capable of very high speeds, but also has
`
`shortcomings. Ex. 2003 (Goodman), col. 2, ll. 20-38. For example, frequency
`
`tuning can only produce a limited range of impedances. Id. (“The dynamic tuning
`
`range [of frequency tuning] is relatively low.”) As shown by the formula below,
`
`impedance (Z) can be changed by altering capacitance (c) or frequency (f):
`
`- 16 -
`
`Page 17 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`𝑍(cid:3404) 12𝜋𝑓𝑐
`
`Ex. 2005 (Chung) col. 2, l. 44. Since available RF sources provide a limited range
`
`of frequency values, they also provide a limited range in impedance values for
`
`impedance matching. Ex. 2004 (Goodman), col. 2, ll. 20-38. For example,
`
`conventional RF sources used in semiconductor fabrications systems can change
`
`their frequency only ±5% or ±10%. See, e.g., Ex. 2006 (AE Generator Data Sheet)
`
`at 2 (stating in the specifications table, under “Frequencies,” frequency tuning
`
`ranges of ±10% and ±5%); Ex. 2007 (MKS Generator Specification) at 3 (stating
`
`in section 3.2.1 a frequency range of ±5%); Ex. 2008 (ENI Generator
`
`Specification) at 7 (stating in section 4.1 a frequency range of ±5%). Zhang states
`
`that “the source [RF generator] may be able to vary frequency within about +/-5
`
`percent ….” See Ex. 1006 (Zhang), col. 3, ll. 28-29.
`
`Using only frequency tuning to impedance match in a semiconductor
`
`processing system would require a very wide frequency range that RF generators
`
`cannot practically provide. Ex. 2004 (Goodman), col. 2, ll. 20-38. As one prior art
`
`patent stated:
`
`A matching network having fixed reactive elements can be used to
`transform a reactive load to a load that appears purely resistive and
`can also be efficiently driven by a variable frequency RF supply. This
`approach, however, would typically require a very wide frequency
`range, e.g. +/-30%, because the load impedance can vary widely,
`- 17 -
`
`
`
`
`Page 18 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`e.g., +/-200%. Such a wide frequency range is unacceptable for
`processing reasons and also because of potential interference with
`other equipment protected using narrow-band filters.
`
`Id. (emphasis added).
`
`Variable capacitors, on the other hand, are not so limited in range, and thus
`
`“can accommodate a widely varying load.” Id., col. 2, ll. 39-40; see also Ex. 1001
`
`(‘122 patent), col. 12, l. 45 to col. 13, l. 7. An EVC can provide a large range of
`
`potential capacitances by simply including amongst its discrete capacitors a small-
`
`value capacitor and a large-value capacitor. See id. Range can also be enhanced
`
`by increasing the number of discrete capacitors. See id. Thus, the matching
`
`network of the challenged claims, by using a shunt and series EVC instead of
`
`frequency tuning, provides not only a speed comparable to frequency tuning, but
`
`also a range that could not be provided by frequency tuning. See id.
`
`III. Claim construction
`Only for purposes of this Preliminary Response, Reno does not dispute
`
`Petitioner’s proposed claim constructions. Reno reserves its rights in all regards to
`
`dispute Petitioner’s proposed claim constructions and proffer its own.
`
`
`
`
`- 18 -
`
`Page 19 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`IV. The Petition’s grounds of supposed rejection identified in the Petition’s
`table are different from the grounds described in the body of the
`Petition and this response uses the nomenclature of the body rather
`than the table.
`On page 4, the Petition provides a table stating six grounds of rejection.
`
`This table, however, labels the grounds of rejection in a manner inconsistent with
`
`the labeling of the grounds in the body of the Petition. For example, the Petition’s
`
`table states that ground 3 is Zhang in view of Chen. Pet. at 4. But the body of the
`
`Petition states that ground 3 is Howald in view of Chen. Pet. at 33. For the sake of
`
`consistency, this Patent Owner Response will label the grounds consistent with the
`
`labeling used in the body of the Petition. Accordingly, the grounds of rejection are
`
`as follows:
`
`Ground
`1
`2
`3
`4
`
`5
`6
`
`35 U.S.C. Prior Art
`102(a)
`Zhang
`103(a)
`Zhang
`103(a)
`Howald in view of Chen
`103(a)
`Howald in view of Chen
`and Bhutta121
`Zhang in view of Chen
`Howald in view of Chen,
`Bhutta121, and Scanlan
`
`103(a)
`103(a)
`
`Claims
`1-4, 6-11
`1-4, 6-11
`1-12
`1-12
`
`4-5, 9, 12
`10-12
`
`V.
`
`
`Summary of Zhang and Chen
`Because the Petition provides almost no description of the cited references,
`
`Patent Owner provides a brief overview of Zhang and Chen.
`
`
`
`
`- 19 -
`
`Page 20 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
` Zhang (Ex. 1006)
`Patent Owner cited Zhang during original prosecution. See Ex. 1001 (‘122
`
`patent) at 2.
`
`1.
`Substrate process system
`Zhang discloses an error feedback-based method of capacitive tuning to
`
`cause an impedance match. Ex. 1006 (Zhang), Abstract. Fig. 1 shows a substrate
`
`process system 100 that includes a substrate process chamber 101, two RF sources
`
`112, 116, and two corresponding matching networks 110, 118. Id., Fig. 1.
`
`Fig. 1 of Zhang
`
`
`
`Zhang’s first matching network can include variable capacitors for enabling
`
`an impedance match. See Id., Fig. 2; col. 4, ll. 30-61. For example, the matching
`
`circuit 210 of Fig. 2 adjusts the capacitance of the shunt variable capacitor C1 to
`
`achieve an impedance match. Id. col. 4, ll. 40-42. The system 100 adjusts the
`
`
`
`
`- 20 -
`
`Page 21 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`capacitance of the shunt variable capacitor C1 to achieve an impedance match
`
`based on the reflected RF power, gradually increasing or decreasing the
`
`capacitance with the goal of minimizing the reflected RF power. See id., col. 4, ll.
`
`1-29; Figs. 3-6.
`
`Reflected RF power is monitored using first and second indicator devices
`
`150, 152. Id., Fig. 1. Zhang further provides:
`
`When reflected power is used as the indicator, the indicator devices
`150 and 152 are respectively coupled between the power sources 112
`and 116 and the matching networks 110 and 118. To produce a signal
`indicative of reflected power, the indicator devices 150 and 152 are
`directional couplers coupled to a RF detector such that the match
`effectiveness indicator signal is a voltage that represents the
`magnitude of the reflected power. A large reflected power is
`indicative of an unmatched situation. The signals produced by the
`indicator devices 150 and 152 are coupled to the controller 114.
`
`Ex. 1006 (Zhang), col. 4, ll. 10-19. Thus, reflected power is obtained by
`
`positioning a directional coupler 150 (or other indicator device) between an RF
`
`source 112 and a matching network 110. Id. The directional coupler 150
`
`generates an indicator signal, which is a voltage representing the magnitude of the
`
`reflected power. Id. This signal is sent to the controller to begin the tuning
`
`process. Id.
`
`- 21 -
`
`Page 22 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`2.
`Error feedback-based tuning methods
`The flowcharts of Figs. 3-6 illustrate the different tuning methods disclosed
`
`by Zhang. These flowcharts show different embodiments of the same essential
`
`tuning process, where tunable elements (e.g., capacitors) are increased (step 308)
`
`and decreased (step 312) to test their effect on the reflected power (steps 310 and
`
`314). Ex. 1006 (Zhang), Fig. 3. The tunable element then moves (increases or
`
`decreases) to the position that results in the lowest reflected power (step 320). Id.
`
`The process 306 then starts again to enable the tunable elements to gradually arrive
`
`at the positions necessary to minimize reflected power and thus achieve an
`
`impedance match. Id.
`
`For example, Fig. 3 illustrates a method 300 of tuning a matching network
`
`according to one embodiment. Id., Fig. 3; col. 5, l. 46 to col. 6, l. 34.
`
`
`
`
`- 22 -
`
`Page 23 of 53
`
`ADVANCED ENERGY INDUSTRIES INC.
`Exhibit 1017
`
`

`

`Case IPR2019-00248
`U.S. Patent No. 9,496,122
`
`Fig. 3 of Zhang
`
`The system first reads an initial reflected RF power (step 302). Id. When the
`
`
`
`reflected power is not within the acceptable range, the tune RF match process 306
`
`begins. Id. The tunable element is increased from an initial setpoint by a
`
`predetermined step size (step 308), and the first adjusted reflect