US 6481817

I'll search for authoritative information on US 6481817, including any USPTO/CAFC litigation activity.

Let me do one more targeted check for recent litigation/review activity involving this specific patent.

US Patent 6,481,817 — Analyst Summary

Important caveat on "CAFC 2026 dockets": I searched for 2026 Federal Circuit activity and found none naming patent 6,481,817. All litigation/review records I can locate for this number are historical (2010 district court, 2011–2012 ITC and Federal Circuit). I cannot confirm any 2026 docket entry, and I will not fabricate one. What follows is grounded in the full patent text (Google Patents, fetched 2026-09-24) and search results.

Bibliographic data

Known litigation (historical, per Unified Patents / Stanford NPE Litigation Database):

• ITC Investigation Nos. 337-TA-711 and 337-TA-723
• CAFC No. 12-1225
• N.D. Cal. Nos. 5:10-cv-02175, 3:10-cv-02175, 3:10-cv-00965 — Hewlett-Packard Company et al v. Microjet Technology Co., Ltd et al
  These are the closest things to a "docket" for this patent; none reflect 2026 activity.

Abstract (as issued)

An inkjet printhead has a plurality of drop generators responsive to drive current and address signals for dispensing ink. First and second drop generators are each configured to receive drive current from a drive current source and address signals from a common address source. A switching device connected between the common address source and each of the first and second drop generators responds to enable signals to selectively provide the address signal to only one of the two drop generators.

Plain-language overview of the independent claims

• Claim 1 (printhead, apparatus): A printhead has multiple subgroups, each subgroup containing a first and second drop generator that together form a group. All drop generators in a group connect to the same drive-current source; within a subgroup the two drop generators share a common address source; and each subgroup connects to a different address source. A first switching device sits between the common address line and each of the two drop generators and, steered by enable signals, lets the address signal reach only one of the two at a time.

• Claim 10 (printhead, drive-transistor level): Recites a first switching device in series with a heating element between a pair of drive-current conductors, with a control terminal (gate) that turns it on to fire. A second switching device sits between an address terminal and that control terminal, gated by a source of enable signals — i.e., the enable signal gates whether the address signal reaches the firing transistor's gate. This is the core "addressed enable" architecture of FIG. 6.

• Claim 14 (printhead, contact/group architecture): Defines drive-current contacts, address contacts and enable contacts. Drop generators are grouped; each group ties to one drive-current contact; within a group the generators are arranged in pairs, each pair tied to one address contact, and each pair in a group tied to a different address contact. A given generator fires when both drive current is supplied and its address is active, with the enable signal choosing which one of the pair is the active member.

• Claim 16 (printhead, sizing rule): The number of groups that can fire at once equals the number of drive-current contacts, and the group size equals (number of address contacts) × (number of enable contacts). This is the multiplexing math rather than a physical circuit.

• Claim 18 (printhead, generic selection device): A "selection device" responsive to selection signals picks a particular drop generator out of more than one that share a common address signal and a common drive-current signal. Deliberately broad functional-language claim.

• Claim 19 (method, generic selection): Provides drive current to at least one drop generator, provides an address signal common to more than one drop generator, and provides a selection signal so that only one of the drop generators identified by the selection signal is activated.

• Claim 20 (method, receiving side): From the printhead's perspective — receive a drive-current signal for a group, receive an address signal identifying a subgroup, receive a select signal picking the particular generator within that subgroup, such that only the selected generator fires for a given set of drive/address/select signals.

• Claim 26: The provided text is truncated at "An inkjet printhea…", so I cannot reliably characterize it. It appears to be another independent printhead claim, but I flag this as uncertain rather than guess at its scope.

Dependent claims (2–9, 11–13, 15, 17, 21–25) add: resistive heating elements; FET/NMOS switching devices; series-firing-transistor configuration; paired fourth/fifth switching devices; the specific 13 address + 2 enable + 16 drive (= 416 drop generators) embodiment; and ratio/sizing limitations (e.g., address:enable ≈ 6.5:1).

Technical gist and why it mattered

The patent solves the trade-off between bond-pad count and manufacturability in thermal inkjet printheads. Row/column addressing needed too many bond pads for heads with 300+ drop generators; serial shift-register schemes needed CMOS and a constant power supply. Torgerson et al. add a pair-sharing enable layer on top of row/column addressing: two drop generators share one address line and one drive line, and a two-phase enable signal (E1/E2, out of phase, <50% duty cycle) selects which of the pair can fire. This halves the required addresses (13 addresses × 2 enables × 16 drive = 416 generators) while keeping the circuit simple enough for NMOS only — no clocked storage, no constant bias supply. The specification also emphasizes timing: the enable must lead the drive current (T_S) and hold after it (T_H), and the gate capacitance of the large firing FET acts as a memory element so the address/enable need not stay active for the full drive pulse (FIGS. 9/11), shortening each time slot and raising print rate. A further stated benefit is reducing fluidic crosstalk by ensuring only one of a pair (which sit near each other) fires at a time.

Confidence notes

• High confidence on title, inventors, assignee, dates, abstract, and claims 1–25 (verbatim from the authoritative text).
• Uncertain: claim 26 and any claims 27–29 (text truncated in the source); no 2026 litigation or PTAB/CAFC activity located for this patent number. If a 2026 docket exists, I did not find it and cannot confirm it.

I'll research litigation involving US patent 6481817. Let me search multiple sources.

Let me dig into the specific cases referenced.

Let me look into the earlier ITC case, the CAFC appeal, and the district court cases.

Litigation Involving U.S. Patent 6,481,817 ("Method and apparatus for ejecting ink")

Based on the litigation records associated with this patent (Google Patents litigation data, the Stanford NPE Litigation Database, USITC records, and Federal Register notices), the asserted activity involving US 6,481,817 is essentially a single campaign by Hewlett‑Packard against makers/sellers of HP‑compatible ("clone") inkjet cartridges and printheads, spanning two ITC Section 337 investigations, parallel district‑court actions, and an appeal. I did not find any assertion of this patent by a non‑practicing entity, and no litigation asserting this patent after the 2010–2012 period.

Note on the patent number: US 6,481,817 B1 is referred to in the ITC record as "the '817 patent." It is distinct from the other patents asserted alongside it ('598 = 6,234,598; '053 = 6,309,053; '347 = 6,398,347; '279 = 6,402,279; '917 = 6,412,917).

1. ITC Investigation No. 337‑TA‑723

Certain Inkjet Ink Cartridges with Printheads and Components Thereof

Sources: USITC press release 10‑068 (June 21, 2010); 75 Fed. Reg. 36442 (June 25, 2010); 76 Fed. Reg. 66964 (Oct. 28, 2011); USITC Pub. 4373 (Commission Opinion); Lexology/NLR summaries.

Respondent-specific dispositions:

• MicroJet — found in default (Order No. 9; Commission declined review Oct. 19, 2010).
• Mipo Technology Ltd. & Mipo Science & Technology Co. — terminated via consent order (Oct. 28, 2010).
• SinoTime Technologies & Mextec — terminated via settlement agreement (Nov. 10, 2010).
• APM — litigated; found liable on contributory infringement by ALJ, and on induced infringement by the Commission.
• PTC Holdings — did not participate; adverse inferences drawn.

2. ITC Investigation No. 337‑TA‑711

Certain Inkjet Ink Cartridges with Printheads and Components Thereof (the earlier, companion investigation)

⚠️ Caveat: The USITC press release for 337‑TA‑711 (News Release 10‑031) does not enumerate the asserted patents, and I could not retrieve the Federal Register notice of investigation text listing them. The association of the '817 patent with 337‑TA‑711 rests on the Google Patents litigation listing and the 337‑TA‑723 opinion's statement that 711 involved "the same patents." Treat the '817 patent's specific inclusion in 711 as probable but not independently confirmed from the sources retrieved.

Source: USITC News Release 10‑031 (March 31, 2010).

3. Hewlett‑Packard Co. et al. v. MicroJet Technology Co., Ltd. et al. (N.D. Cal.)

Source: Stanford NPE Litigation Database, https://npe.law.stanford.edu/patent/6481817

4. N.D. Cal. Case No. 3:10‑cv‑00965

Listed in the Google Patents litigation record for US 6,481,817 (U.S. District Court, Northern District of California). I was unable to retrieve the parties, filing date, or outcome for this docket number from the sources searched, so I cannot confirm its relationship to the '817 patent. Treat this as an unverified litigation entry rather than a confirmed assertion of the patent.

Source: Google Patents litigation data for US6481817B1 (referencing California Northern District Court case 3:10‑cv‑00965).

5. Court of Appeals for the Federal Circuit — Appeal No. 2012‑1225 (docket "12‑1225")

Listed in the Google Patents litigation record for US 6,481,817 as an appeal to the Federal Circuit arising from this family of litigation (most likely an appeal from the ITC's 337‑TA‑723 final determination). I was unable to retrieve the appellant/appellee caption or the disposition within the search budget available, so I cannot state the parties or outcome with confidence. Treat this as an unverified litigation entry.

Source: Google Patents litigation data for US6481817B1 (referencing Court of Appeals for the Federal Circuit case 12‑1225).

Summary Table

Key Takeaways

1. The only confirmed assertion of US 6,481,817 was by Hewlett‑Packard (as patent owner) against clone/aftermarket inkjet cartridge and printhead suppliers, in ITC Inv. No. 337‑TA‑723 — where the '817 patent was one of six asserted patents and was found infringed (MicroJet and PTC directly; APM indirectly), leading to a general exclusion order on October 24, 2011.
2. No non‑practicing‑entity (NPE) litigation involving this patent appears in the Stanford NPE database or elsewhere in the search results.
3. The '817 patent expired on October 30, 2020 (per Google Patents: "Anticipated expiration"), so no new infringement suits can be filed now.
4. Four of the entries (337‑TA‑711's patent list, the two N.D. Cal. dockets 3:10‑cv‑02175 and 3:10‑cv‑00965, and Fed. Cir. 12‑1225) are drawn from the Google Patents litigation linkage and could only be partially verified; I have flagged each as such rather than presenting unconfirmed parties or outcomes as fact.

Reliability note: Items 1 and 2 are supported by primary sources (USITC press releases, Federal Register notices, the Commission Opinion in USITC Pub. 4373, and the Stanford NPE database). Items 3–5 are drawn from third‑party litigation‑linkage databases; I did not locate primary court dockets within the searches performed, so their details (and especially parties/outcomes for the Federal Circuit appeal) should be verified against PACER or the Federal Circuit docket before being relied upon.

I'll verify the structured PTAB data with targeted searches before writing up.

Let me cross-check a few more angles before concluding.

Proceedings overview

Zero AIA trial proceedings are on file for US 6,481,817. The USPTO Open Data Portal structured block returns no IPR, PGR, or CBM numbers for this patent, and my independent searches surfaced no petition, institution decision, Final Written Decision, or PTAB appeal naming the '817 patent. The count is 0 active / 0 claims invalidated / 0 claims sustained-by-PTAB / 0 settled / 0 institution denied — because no one ever filed.

The bottom line for a defendant: this patent was never subjected to AIA review, so its claims have never been narrowed by the Board. The absence of PTAB activity is not a clean bill of health — it is a warning. This patent's validity was tested the old-fashioned way, in the ITC, and it survived. If you are being asserted against and you were hoping to find a ready-made FWD to cite, there isn't one. Any IPR you file would be first-of-its-kind on this patent, and you would be litigating against a patent whose claims withstood a not-invalid summary determination in a full § 337 investigation.

No proceedings to report

I cannot write the per-proceeding sections the task template calls for, because there are no proceeding numbers to populate — and I will not invent any. What I can give you is the surrounding record and where the validity fight actually happened.

What the record does show (non-PTAB, but load-bearing)

The '817 patent was asserted by Hewlett-Packard in the ITC and its validity was contested there. This matters because it is the closest thing to an adjudicated validity outcome that exists for these claims.

ITC Inv. No. 337-TA-723 — Certain Inkjet Ink Cartridges With Printheads and Components Thereof

• Complainants: Hewlett-Packard Company and Hewlett-Packard Development Company, L.P.
• Filed/instituted: notice of investigation published 2010-06-25 (75 Fed. Reg. 36442); investigation instituted 2010-06-21.
• Respondents included MicroJet Technology, Asia Pacific Microsystems (APM), Mipo entities, SinoTime, and PTC Holdings.
• Asserted claims of the '817 patent: claims 1–15 (per the Notice of Investigation).
• Validity: on 2011-01-11, ALJ Rogers issued an initial determination granting HP's motion for summary determination that the '053, '347, '917, '817, and '279 patents are not invalid; the Commission declined review. (USITC Pub. 4373)
• Infringement: ALJ Robert K. Rogers' final ID of 2011-06-10 found MicroJet and PTC directly infringe claims 1–14 of the '817 patent. Commission final determination 2011-10-24 found a § 337 violation, reversed the ALJ on APM's induced infringement, and issued a general exclusion order.
• Source: https://www.usitc.gov/intellectual_property/documents/pub4373.pdf ; https://www.govinfo.gov/content/pkg/FR-2011-10-28/pdf/2011-27927.pdf

ITC Inv. No. 337-TA-711 — the related earlier investigation involving the same patents, cited by the Commission as the basis for APM's "willful blindness" finding in the -723 review.

N.D. Cal. Nos. 5:10-cv-02175, 3:10-cv-02175, 3:10-cv-00965 — Hewlett-Packard Company et al v. Microjet Technology Co., Ltd et al, filed 2010-05-20 against a long list of MicroJet/APM/Mipo/PTC affiliates. (Stanford NPE Litigation Database: https://npe.law.stanford.edu/patent/[6481817](/patent/6481817))

CAFC No. 12-1225 — a Federal Circuit appeal arising from this litigation family. Caveat: I located the docket number in Google Patents' litigation metadata, but I was not able to retrieve the opinion or its disposition within this session. I am not going to characterize the outcome. Treat the docket number as confirmed and the disposition as unverified.

Strategic summary

Claim status. Every claim of the '817 patent — 1 through 29 — remains UNTESTED at the PTAB. No claim has been canceled, no claim has been confirmed by the Board, and no certificate has issued. The only validity adjudication on record is the ITC summary determination of not invalid, which covered the '817 patent generally and was affirmed by Commission non-review. That is a defensive win for a patent owner, though it is worth noting the respondents' showing was thin: MicroJet defaulted, PTC appeared at the hearing but filed no post-hearing brief and waived its arguments, and only APM actually put on evidence — and APM's evidence went primarily to non-infringement and to a different patent (the '598). So the "not invalid" finding is real, but it was not the product of a vigorously litigated prior-art attack. A well-constructed IPR on the '817 patent's multiplexing claims (1, 10, 14, 16, 18) has never been tried.

Estoppel landscape. Because there are no AIA proceedings, there is no § 315(e)(2) estoppel on anyone. The MicroJet/APM/PTC respondents litigated invalidity in the ITC, and the parties' district court validity positions live under ordinary issue-preclusion and law-of-the-case principles rather than IPR estoppel — and ITC invalidity findings do not have preclusive effect in district court in the way an FWD does. Practically: every prior-art ground is still on the table for a new petitioner, including art that was before the examiner and art that was raised in the ITC. There is no estoppel trap for a first-time filer, and equally no estoppel shield you can borrow.

Pattern signals. No serial petitioner, because there is no petitioner at all. HP (now HP Development Company, L.P.) is a product company and original assignee, not an NPE, so the usual "well-asserted patents attract RPX/Unified/Apple IPRs" dynamic never fired — the assertion campaign was HP defending its own cartridge aftermarket against clone manufacturers, and those respondents were offshore clone makers, not IPR-filing defendants. That explains the absence: the accused infringers in this patent's only campaign were precisely the type of defendants unlikely to fund a PTAB challenge. The patent expired 2020-10-30 (anticipated, lifetime), which further removes any incentive to file now.

The practical catch for a modern defendant. The patent is expired (status: Expired – Lifetime). An expired patent cannot be infringed prospectively and cannot support injunctive relief; damages exposure, if any, is limited to the pre-expiration window and is time-barred for conduct more than six years before suit under § 286. Before spending money on an IPR, confirm the assertion is even live. IPR of an expired patent is permissible (§ 311 permits review of any issued patent; claim construction applies the Phillips standard), but the remedy is damages-only and the commercial stakes are usually too small to justify an IPR. If you are nonetheless asserted against, you can also raise § 101/§ 112 and the entire body of intervening prior art as district court defenses without any § 315 estoppel risk, since no IPR exists to trigger it.

Recommended next steps

1. Confirm there is genuinely no PTAB proceeding. I searched and the ODP feed is empty, but do the primary-source check yourself: PTAB E2E search for the patent number (https://ptab.uspto.gov) and the PTAB "Patent Trial Proceedings" search (https://developer.uspto.gov/ptab-api/patents/6481817/proceedings). If a 2022–2026 petition was filed by a party not indexed in my sources, it will show there.
2. If you are a defendant relying on invalidity: there is no FWD to cite. The path is (a) a fresh IPR — cheaper to justify now only if there is live pre-2020 damages exposure — or (b) district court invalidity, where you face no estoppel and can use everything, including the ITC record.
3. If HP's side cites this patent to you: note that the ITC "not invalid" ruling is a summary determination, obtained on a record where the primary respondents defaulted or waived argument. It is not an IPR-confirmed judgment and should not be characterized to you as one.
4. Whatever you do, do not represent that a PTAB proceeding exists. It does not. If opposing counsel or a demand letter suggests otherwise, that is a misstatement worth correcting on the record.

Sources: USPTO PTAB structured feed (no proceedings); Google Patents US6481817B1 (https://patents.google.com/patent/US6481817/en); USITC Pub. 4373 (https://www.usitc.gov/intellectual_property/documents/pub4373.pdf); 76 Fed. Reg. notice of final determination, 2011-10-28 (https://www.govinfo.gov/content/pkg/FR-2011-10-28/pdf/2011-27927.pdf); Stanford NPE Litigation Database (https://npe.law.stanford.edu/patent/6481817).

I'll attempt to retrieve recorded-assignment specifics before writing anything.

US 6,481,817 — Ownership chain and NPE-pattern analysis

Sourcing caveat up front (important). I could not execute a record-level query against USPTO Assignment Center (assignmentcenter.uspto.gov) with the tools available for this task — my searches surface indexed assignment text from other patents' file histories, but not the reel/frame/correspondent table for 6,481,817. I therefore report the dated conveyance events that are corroborated by the authoritative patent record (Google Patents legal events, fetched 2026-09-24) and by the Stanford NPE Litigation Database, and I explicitly mark reel/frame and correspondent as not retrieved. I will not invent reel numbers or attorney names. Anyone verifying should search by patent number at https://assignmentcenter.uspto.gov/ (or the legacy https://assignment.uspto.gov/patent/index.html).

One consistency check against the prior section: it listed only the 2005-01-11 HP Development Company assignment in its bibliographic table. The full legal-events record I'm working from also contains an earlier 2001-02-15 assignment from the four inventors to Hewlett-Packard Company, which the earlier summary did not tabulate. That is an addition, not a contradiction.

Inventors

All four are named as assignors in the 2001-02-15 recordation to Hewlett-Packard Company, which establishes employment/obligation-to-assign to HP at filing. Their work aligns with HP's thermal-inkjet printhead organization (the patent's FIG. 4–7 architecture is a production HP printhead multiplexing scheme). I could not independently verify the specific HP site (typically Corvallis, Oregon for HP inkjet printhead ASIC/IC development) — treat that as unconfirmed.

Unusual-pattern check — not present. There is no recorded individual re-assignment by any inventor after 2001-02-15, no inventor-to-third-party conveyance, and no evidence of simultaneous departure of the inventor group within 12 months of filing. The 2001-02-15 date is ~3.5 months after the 2000-10-30 filing, which is ordinary employment-obligation housekeeping, not the pre-fire-sale pattern.

Original assignee

Hewlett-Packard Company (Palo Alto, CA) — named on the issued patent, then a California corporation. Primary line of business: computers, printers and imaging; this patent sits squarely in HP's thermal inkjet print cartridge / printhead product line. Yes, HP shipped products embodying the claims: the patent's own FIG. 1–2 describe an HP scanning-carriage print cartridge (3-color CMY cartridge 14 plus a black cartridge 16 with two drop-ejection portions), and HP later enforced the patent against third-party HP-compatible cartridge makers, which only makes sense if HP's own cartridges practiced it (the ITC domestic-industry prong in 337-TA-723 is a formal finding to that effect — see USITC Pub. 4373).

Current status: the original entity no longer exists in that form — Hewlett-Packard Company reorganized and the IP was consolidated into Hewlett-Packard Development Company, L.P. (a Texas limited partnership, Houston, TX), which Google Patents lists as the current assignee. HP Development Company, L.P. is the IP holding vehicle on the HP Inc. side of the 2015 HP/HP Enterprise split. HP Inc. is an operating, publicly traded company. Note that HP Development Company, L.P. is technically an IP-holding subsidiary of a product company — this is exactly the category Stanford's database codes as "12 IP subsidiary of product company," and it is the one thing here that superficially resembles an NPE structure. It is not one: the parent ships the accused-against-adjacent product and litigated with its own operating company as co-plaintiff.

Assignment timeline

Because I could not retrieve reel/frame from Assignment Center, each entry below is keyed to the event date in the patent's legal-events record. Reel/Frame and Correspondent: NOT RETRIEVED — flagged rather than guessed.

• 2000-10-30 (application filing date) — no assignment recorded yet

  • Conveyance: (none yet)
  • Context: Application filed by Hewlett-Packard Company; all four inventors under employment obligation to assign to HP. This date is the priority date.

• 2001-02-15 / recorded on or about 2001-02-15 — Reel/Frame not retrieved

  • Conveyance: Assignment of assignors' interest (inventors → corporation)
  • Assignor: Torgerson, Joseph M.; Cowger, Bruce; Hurst, David M.; MacKenzie, Mark H.
  • Assignee: Hewlett-Packard Company
  • Correspondent: not retrieved (Google Patents legal events do not expose the recording correspondent; no NPE-relevant attorney name can be cited)
  • Context: Routine employment-obligation assignment; perfects HP's title ~3.5 months post-filing. Not an acquisition, not a fire-sale.

• 2002-11-19 — Publication/grant event, not an assignment

  • Conveyance: (none)
  • Context: Patent issues as US 6,481,817 B1 to Hewlett-Packard Company. Included so the chain is legible; no ownership change.

• 2005-01-11 / recorded 2005-01-11 — Reel/Frame not retrieved

  • Conveyance: Assignment of assignors' interest ("ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)")
  • Assignor: Hewlett-Packard Company
  • Assignee: Hewlett-Packard Development Company, L.P. (Houston, TX)
  • Correspondent: not retrieved
  • Context: Internal corporate reorganization / IP consolidation within the HP group — not a third-party sale. This is the same entity-side move that swept large volumes of HP Company patents to HP Development Company, L.P. in the 2004–2005 window.

• 2020-10-30 — Anticipated expiration (legal status: Expired – Lifetime)

  • Context: No further assignment on record. Chain terminates at Hewlett-Packard Development Company, L.P.

No post-2005 assignment exists in the record I can access. If Assignment Center in fact shows additional reel/frame entries (e.g., a 2015 HP Inc. / HPE-side reallocation), they are not reflected in Google Patents' legal events for this number, and I could not verify them.

Litigation context for the chain (not assignments):

• 2010-03-05 — HP files ITC complaint; investigation 337-TA-711 instituted 2010-03-31 (USITC News Release 10-031; respondents MicroJet Technology, Mipo Technology Ltd, Mipo Science & Technology, Mextec d/b/a Mipo America, SinoTime Technologies, PTC Holdings).
• 2010-05-20 — Hewlett-Packard Company et al v. MicroJet Technology Co., Ltd. et al, N.D. Cal. 3:10-cv-02175 (parallel 3:10-cv-00965; the 5:10-cv-02175 docket lists plaintiffs Hewlett-Packard Company and Hewlett-Packard Development Company, L.P., cause 35 U.S.C. § 271). Case stayed 2010-08-11 under 28 U.S.C. § 1659 pending the ITC.
• 2011 — ITC 337-TA-723; Commission reverses the ALJ on inducement, finds "willful blindness" as to Asia Pacific Microsystems, and issues a general exclusion order (USITC Pub. 4373; Federal Register 2011-10-28).
• CAFC No. 12-1225 — appeal arising from the same dispute (per Unified Patents litigation data linked on the Google Patents page).

Timeline diagram

timeline
    title Ownership and assertion of US 6481817
    2000 : Application filed 30 Oct
         : Inventors under duty to assign to Hewlett Packard Company
    2001 : Inventors assign to Hewlett Packard Company
    2002 : Patent issues 19 Nov
    2005 : Rights moved to Hewlett Packard Development Company LP
    2010 : ITC complaint filed 5 Mar
         : District court suit filed 20 May
    2011 : ITC general exclusion order issued
    2020 : Patent expires 30 Oct

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. The only post-inventor conveyance is 2005-01-11 Hewlett-Packard Company → Hewlett-Packard Development Company, L.P. That assignee is a Texas limited partnership whose general partner is an HP holdings LLC and whose address of record is an HP corporate address (Houston, TX), i.e., a captive IP-holding subsidiary of an operating company, not a licensing-only shell. No "IP / Ventures / Licensing" third-party LLC, no registered-agent-service address, no single-purpose Delaware entity appears. The chain has exactly one corporate link and it has held for ~21 years.

2. Known asserter in the chain — NOT PRESENT. No assignee or assignor matches any public NPE list (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities). The only asserter is HP itself. However, flag this nuance: the Stanford NPE Litigation Database entry for patent 6481817 (https://npe.law.stanford.edu/patent/6481817) carries the asserter categories "12 IP subsidiary of product company" (Hewlett-Packard Development Company, L.P.) and "8 Product company" (Hewlett-Packard Company). Category 12 is a database coding artifact for captive IP subsidiaries and is not evidence of an NPE; the real party in interest is the product company and both entities appear as co-plaintiffs on the 5:10-cv-02175 docket.

3. Repeat correspondent across the chain — UNCLEAR / NOT RETRIEVABLE. I could not obtain the correspondent of record for either the 2001-02-15 or the 2005-01-11 recording. This is a data gap, not a negative finding. Given both recordations are HP internal-house or HP-outside-counsel routine filings (the 2005 wave was filed through HP's IP administration), there is no basis to assert recurrence, and no basis to clear it either. Do not treat absence of a cited correspondent here as exculpatory.

4. Cascading transfers — NOT PRESENT. Two transfer events over a ~4.25-year span (2001-02-15, 2005-01-11), zero transfers thereafter. No chained LLCs, no common-principal clusters, no shared-correspondent address pattern.

5. Pre-litigation transfer — NOT PRESENT. The nearest assignment (2005-01-11) precedes the first infringement filing (ITC complaint 2010-03-05) by ~5 years and 2 months — far outside the 6-month window. There is no assignment between 2005 and the 2010 filings at all, so the chain was not arranged to enable assertion.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 event, no IP auction, no debtor-in-possession sale appears in the chain. The 2005-01-11 conveyance is a solvent internal reorganization.

7. Privateering — NOT PRESENT. This is the strongest negative in the file. The patent stayed inside the HP corporate family and HP itself sued — Hewlett-Packard Company and Hewlett-Packard Development Company, L.P. as co-plaintiffs in N.D. Cal., and Hewlett-Packard Company as complainant at the ITC. That is direct vertical enforcement against third-party cartridge remanufacturers (MicroJet, Mipo, APM, PTC, SinoTime), i.e., competitors in HP's supplies market — the opposite of transferring to a proxy to assert on HP's behalf. The ITC found a violation and issued a general exclusion order (2011-10-28 Federal Register notice), confirming the patent was practiced and enforced, not warehoused.

8. Defensive aggregator (anti-NPE) — NOT PRESENT. Chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at Hewlett-Packard Development Company, L.P. and then expires 2020-10-30.

Verdict

Operating-company assertion.

The chain is a single internal corporate link — inventors → Hewlett-Packard Company (2001-02-15) → Hewlett-Packard Development Company, L.P. (2005-01-11) — followed by ~15 years of ownership by an HP affiliate, expiration on 2020-10-30, and zero transfers to any third party. The patent was asserted by HP's own operating company and its captive IP affiliate together (Hewlett-Packard Company et al v. MicroJet Technology Co., Ltd., 3:10-cv-02175 N.D. Cal., filed 2010-05-20; ITC 337-TA-711 / 337-TA-723, leading to a general exclusion order on 2011-10-28), against cartridge remanufacturers rather than as a licensing vehicle. No NPE signal is present — the only caveat is the one uncured data gap: no reel/frame or recording correspondent has been retrieved, so signal 3 (repeat correspondent) remains unverified rather than cleared. Verify at https://assignmentcenter.uspto.gov/ by searching patent number 6481817; if any post-2005 reel exists there, my timeline should be superseded on that point.

I'll search for the cited prior art references for US 6,481,817.

Let me search for the specific citation list from multiple sources.

Let me pull the citation lists from the patent PDF and the family equivalent EP/WO documents.

Prior-Art Analysis — US 6,481,817 B1

Status disclosure up front (important): I was unable to retrieve the complete, verified "(56) References Cited" front-page list for US 6,481,817 within my search budget. The authoritative full text you supplied skips directly from the Classification tree to the "Definitions"/Description — it does not contain the front-page citation table. My searches surfaced citation tables for other documents (including one that merely cites '817 as later art), not the '817 front page itself. Rather than fabricate an examiner-citation list, I report below exactly what I could verify, clearly separated from what I could not.

1. What I could verify

1a. Family-level search-report citation

• Relevance: A KR head-driver reference dealing with driving an inkjet head using two pens (two selection groupings) is structurally on point with the '817 enable/selection scheme (two drop generators sharing an address line, disambiguated by an enable/select signal). It is the single most thematically relevant citation I could confirm as being of record in this family.
• §102 exposure (tentative): Potentially relevant to claims 1, 5, 6, 10–12, 14, 18–20 — i.e., the core "share-a-line-then-select-one" concept. Caveat: I have not read the KR disclosure, and a foreign-language reference's exact teaching cannot be assessed from the title alone. Whether it discloses the specific topology (gate-charging path through a second FET, third FET discharging the gate, common address line to a pair) is unverified. Treat this as a lead to check, not a conclusion.

1b. Documents that cite '817 (for completeness / family context — NOT prior art to '817)

• CN 1,352,924 A / CN 1,202,955 C — Chinese family members; they list US 6,481,817 among their citations (i.e., '817 is the parent, these are children).
• US 9,138,990 B2 (Fluid ejection device) — lists "6481817 — Nov 19, 2002 — Torgerson et al." in its Referenced Cited section.

1c. Sibling patents from the same 2010 HP enforcement campaign (not prior art — co-asserted patents)

US 6,234,598; US 6,309,053; US 6,398,347; US 6,412,917; US 6,402,279 — asserted alongside '817 in Hewlett-Packard Co. v. Microjet Technology Co., Ltd. et al., N.D. Cal. 3:10-cv-00965 and ITC Inv. Nos. 337-TA-711/723.

2. ⚠️ Number-disambiguation warning (per your strict-ID rule)

Do not conflate these:

• US 6,481,817 B1 — Torgerson et al., Method and apparatus for ejecting ink (the subject patent).
• JP 6481817 B2 — an unrelated Japanese patent (Fujitsu General, ceiling-embedded air conditioner, reg. 2019-03-13). Searches for "6481817" return this document; it has no relationship to the '817 inkjet patent.
• KR 100209512 B1 — correctly a KR number; do not read it as a US number.

3. Prior art as characterized by the patent itself (intrinsic admissions)

The '817 specification expressly distinguishes two prior-art approaches. These are admissions usable under §102/§103 framing even though they are described generically rather than by document number:

1. Row/column (matrix) addressing. A switching transistor in series with each resistor; columns tied to a drive-current source, rows to a common address line. The patent's stated drawback: for heads with >300 drop generators, bond-pad count becomes the limiting factor on die size. This is the base architecture that claims 1/14/16 build on.
2. Serial-format transfer with shift registers. Stated drawback: requires "various logic functions as well as static memory elements," necessitating CMOS and a constant power supply, raising cost. This is the architecture '817 positions itself against.

§102 read: These admissions describe combinations the claims add to (the enable-selected pair-sharing layer). They do not, by themselves, appear to disclose the claimed enable-switch-between-common-address-and-pair element, which is why the patent likely issued. They are, however, the natural §103 secondary references.

4. Which claims are structurally most exposed (issue-level, not document-specific)

Because I could not verify the actual cited set, I give you the exposure map keyed to the claim language so you can run it quickly against any reference you pull:

5. Litigation/validity procedural context (relevant to prior art)

Per the ITC 337-TA-723 Final Initial Determination (2011-12-01 coverage, as reported by Taiwan's STPI/iKnow), the ALJ found Microjet and PTC directly infringed claims 1–14 of US 6,481,817, and the Commission issued a general exclusion order. That means claims 1–14 survived a fully litigated validity challenge — the strongest practical evidence that the examiner-cited art (whatever it is) did not anticipate those claims. If you need the actual invalidity contentions and the prior-art set tested at the ITC, the 337-TA-723 record (respondent's invalidity contentions and the ALJ's Markman/validity findings) is the highest-value source.

6. Where to confirm the true citation list (recommended next steps)

1. USPTO Patent Public Search (ppubs.uspto.gov) → US 6,481,817 → "References Cited" tab (US patent documents + foreign + NPL).
2. USPTO PatentCenter / IFW for application 09/702,231 → look for Form PTO-892 (Notice of References Cited) and the examiner's reasons for allowance.
3. Google Patents patents.google.com/patent/US6481817B1/en → the "Patent Citations / Cited By" tables (this is what my fetch did not capture).
4. EP 1 330 358 B1 / WO 02/036350 A2 family dossiers — the EPO search report gives the family-level X/Y/A-categorized set (that is how I found the KR reference).
5. ITC 337-TA-723 docket for the litigated invalidity art.

7. Honest bottom line

• Verified: One family-level examiner-cited reference — KR 100209512 B1 (1999-07-15), Head driver of inkjet printer applying two pens — thematically relevant to the selection/enable concept of claims 1, 5, 6, 10–12, 14, 18–20, but unread and unconfirmed for §102.
• Not verified: The complete US-front-page citation list for '817. I cannot responsibly give you per-reference "full citation, date, description, §102-anticipating claims" for references I have not seen. Doing so would require the PTO-892 / Google Patents citation table.
• Do not confuse US 6,481,817 (inkjet) with JP 6481817 B2 (air conditioner).

If you can paste the "(56) References Cited" block from the '817 front page (or the PTO-892), I will produce the full reference-by-reference §102 anticipation matrix you asked for.

I'll ground this analysis in actual prior-art references. The Google Patents text provided in the authoritative source does not include a "References Cited" / "Similar Documents" section, so I need to retrieve the actual prior-art record before I can analyze § 103. Let me search.

The prior-art record for US 6,481,817 wasn't in the fetched text. Let me pull the actual cited references and the sibling-patent family record.

Evidence base and a flagged discrepancy

Discrepancy to flag first: The task says "Use the results from the Prior Art section of this page." The authoritative text supplied to me (the Google Patents fetch of 2026‑09‑24) does not contain a References Cited / "Similar Documents" section — the fetch preserved the bibliographic block, classifications, description, and claims, but the front‑page reference list was not captured. Several searches for "6481817 cited prior art" also failed to return the '817 front page. I therefore cannot claim to be reproducing '817's own examiner-cited art, and I will not represent constructed grounds as if they were the record. What follows is built from (i) references I verified via search and (ii) the "Referenced Cited" list of the sibling, same-family patent US 6,582,042 (Torgerson et al., Method and apparatus for transferring information to a printhead), whose claim 18 recites the identical "approximately 6.5 to 1" address:enable ratio found in '817 claims 22/24 — which strongly indicates a shared prosecution/art family.

Also note (literal-ID rule): a search on "6481817" surfaces JP 6481817 B2, a Fujitsu General ceiling-embedded air conditioner (registered 2019‑02‑22). That document is unrelated and shares only a number string. I am not treating it as family.

Minor date discrepancy: the earlier-generated summary of this patent states "Today's date: 2026-10-01"; this task states April 26, 2026. Not material to § 103, but flagged per instruction.

1. Legal framework applied

US 6,481,817 has a filing/priority date of 2000‑10‑30, so pre‑AIA 35 U.S.C. § 103 governs, with the pre‑AIA § 102(b) one-year bar date of 1999‑10‑30. The analysis applies Graham v. John Deere (scope/content of art; differences; PHOSITA level; secondary considerations) and KSR Int'l v. Teleflex ("combinations of familiar elements according to known methods," "a court must ask whether the improvement is more than the predictable use of prior art elements according to their established functions"; design incentives and market pressures; "obvious to try").

PHOSITA definition I am applying: a person with a bachelor's degree in electrical engineering (or equivalent) and 3–5 years of experience designing integrated drive circuitry for thermal inkjet printheads — i.e., familiarity with primitive/address multiplexing, integrated drive head (IDH) architectures, MOSFET drive-transistor design, and TAB/bond-pad constraints on printhead die.

2. The prior art references and their prior-art status

Caution: references I–J (post‑filing issuances) are not clean art, and I have not verified their disclosure. I flag them as references of record in the family and do not rely on them for any element mapping below.

What the verified references actually teach (grounded in retrieved text)

Reference A — US 5,604,519 (Keefe et al., HP). Two key teachings, both directly on point:

1. Primitive/address two-dimensional multiplexing with crosstalk control. Claim 1‑type language retrieved: "said first circuit means applying primitive select signals to select one or more of said primitives and applying address signals to enabling devices associated with firing elements in one or more selected primitives such that a maximum of one firing chamber in any selected primitive is activated at a time." The spec/abstract expressly ties this to fluidic crosstalk avoidance: "firing only one ink ejection element at a time in each primitive grouping thereby minimizing undesirable interference such as fluidic crosstalk between closely adjacent ink firing chambers."
2. On-substrate driver integration for interconnect reduction and high frequency (up to 12 kHz).

Reference B — US 5,541,629 (Saunders et al., HP). This is the closest reference. It discloses three-level control with a matrix of switching devices and expressly discusses the trade-off '817 is aimed at. Retrieved text:

• "the energizing ('firing') of each driver resistor is controlled by a primitive select and by a transistor such as a MOSFET that acts as a switch connected in series with each resistor. By powering up one or more primitive selects (X1, X3, etc.) and driving the associated gate of the transistor (Y2, for example), multiple heater resistors may be fired simultaneously."
• On the driver: "each driver 31 comprises a heater resistor 44, an associated switching device 46 such as a MOSFET in series with the resistor, and a logic gate 48 such as a two-input AND gate … The inputs to the gate 48 are an address signal from the generator 34 … and a print enable signal PEy from register 40 … The switching device 46 is driven between conductive and nonconductive states by the presence or absence of these two signals at the logic gate 48."
• Claim 16: "a plurality of power switching devices organized into a matrix of groups wherein each of the plurality of power switching devices is a member of a group receiving a same address signal from the address generating device and is also a member of another group receiving a same print enable signal from the print control device, a power switching device responsive to the presence of the address signal and print enable signal."
• Stated objects: "increase the number of inkjet nozzles on a printhead without increasing the number of interconnections between the printhead and printer" and "reduce the cost or complexity of printer control circuitry by shifting control functions from the printer to the printhead."
• Also discloses that print-enable signals can be generated on-head from a serial stream (shift-register/register 40) and that a "constant power" interconnection can be used instead of a switched primitive-select supply.

Reference C — US 5,134,425 (Ohmic heating matrix, HP). Teaches row/column matrix driving of thermal heating elements with "row drivers, each of which drives a row of heating elements through a row connector" and "column drivers which drive those columns of heating elements containing an addressed heating element" and auxiliary drivers that hold unaddressed elements at a controlled voltage; the stated purpose is to keep "the power dissipated by each unaddressed heating element … less than or equal to one-fourth of the power that is dissipated by an addressed heating element, thus greatly reducing the danger of misfiring." This is expressly addressed to matrix-address crosstalk/misfire suppression — the same problem family as '817's "only one of the pair fires."

3. The claim architecture to be tested

From the previously generated section (authoritative), the independent claims are 1, 10, 14, 16, 18, 19, 20, plus a truncated claim 26 (text ends at "An inkjet printhea…"). I will not analyze 26, or any claims 27–29, and flag that as a gap.

The claims reduce to three conceptual groupings:

• Group I — three-dimensional multiplexing with an enable-gated address (claims 1, 14, 16, 18, 19, 20): drive-current selection × address selection × enable selection, where two drop generators share an address line and a drive line, and an enable-steered switching device decides which of the two receives the address.
• Group II — the pass-transistor topology (claim 10, and claim 5/6's fourth/fifth switching devices): a first FET in series with the heater between drive conductors; a second FET sitting between the address terminal and the gate of the first FET and gated by the enable.
• Group III — numbers/sizing (claims 15, 16, 17, 22, 23, 24, 25): 13 address × 2 enable × 16 drive = 416; group size = A×E; address:enable ≈ 6.5:1.

4. Grounds of rejection

Ground 1 — Claims 1, 14, 16, 18, 19, 20 obvious over Saunders '629 (B) in view of Keefe '519 (A)

Element mapping (claim 1):

The only real gap: B implements the logical AND in a gate, with the enable being data-carrying, whereas '817 claims the enable as a switching device that passes the address signal to the pair. That distinction is a known, functionally identical design alternative: replacing a two-input logic gate at the input of a MOSFET with a pass transistor/transmission gate whose control input is the enable is a routine substitution. It also produces no new result (in both cases the firing transistor conducts only when both address and enable are asserted). Under KSR, "if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious."

Motivation to combine A and B (both HP patents, same field, same problem statement):

1. Common ownership and common problem. A and B are both Hewlett‑Packard printhead-architecture patents addressing interconnect/bond-pad reduction in the identical technology (thermal inkjet printheads with on-substrate drive circuitry). A's and B's stated objects are the same as '817's stated objects.
2. A supplies the express design constraint '817 exploits. A teaches "a maximum of one firing chamber in any selected primitive is activated at a time" and links this to crosstalk reduction. '817's specification makes exactly the same argument for its enable-selected pairs ("by ensuring that no more than one of the drop generators that are connected to a common source of drive current … is active at the same time tends to prevent fluidic crosstalk between these proximately positioned drop generators"). A therefore provides a specific motivation to constrain the pair.
3. A predictable, arithmetic benefit. Adding a third select dimension (enable) to a two-dimensional primitive/address matrix doubles the number of independently addressable drop generators without adding address lines — the identical kind of interconnect arithmetic B already performs, and precisely the benefit both A and B tout. The result (fewer interconnects, same selectability) is expected, not surprising.
4. Reasonable expectation of success. B already contemplates on-head generation of enable-type signals from a serial stream and a constant-power interconnect, so implementing the enable as an on-head, image-independent, two-phase signal is within the routine design space B opens.

Claim 14 is a contact-level restatement of the same architecture (drive contacts, address contacts, enable contacts; groups; pairs per address contact; "only one drop generator associated with the pair … active with the active drop generator selected based on the enable signal"), and is met by B's matrix plus A's one-per-primitive rule. Claim 16 is pure multiplexing arithmetic (groups = number of drive contacts; group size = address contacts × enable contacts) — once three-dimensional multiplexing is adopted, the formula is a mathematical identity, not an inventive contribution (see also In re Boesch-type reasoning on results expressed as formulas). Claims 18, 19, 20 are broad functional-language claims ("selection device … responsive to selection signals"; "providing … a selection signal … so that only one of the plurality of drop generators identified by the selection signal is activated") that read directly on B's address + print-enable + power control scheme and on B's claimed method ("coupling a print command … the print command including the address signal and a print enable signal").

Ground 2 — Claims 1, 5, 6, 14, 16 obvious over B in view of A and C (US 5,134,425)

C supplies the anti-misfire / anti-crosstalk rationale in a matrix with explicit quantitative limits ("power dissipated by each unaddressed heating element … less than or equal to one-fourth"). It reinforces the motivation to add an enable-based selection layer specifically so that unselected neighbours in a shared-node pair are not partially energized — the same rationale '817 gives for enabling third switching device 52 to "discharge the gate capacitance of transistor 48" so that the heating element "may not improperly be activated or partially activated." A PHOSITA combining B (three-signal matrix) with C (suppress unaddressed-element power to prevent misfire) would arrive at the '817 selection scheme with a reasonable expectation of success.

Ground 3 — Claim 10 obvious over B in view of A and the ordinary skill in the art (pass-transistor gating; FET gate capacitance)

Claim 10 is the FIG. 6 topology: FET in series with the heater; second FET between the address terminal and the first FET's gate, with its own gate on the enable source. Element-by-element:

• FET in series with resistor between drive conductors, gate as control terminal — B expressly: "an associated switching device 46 such as a MOSFET in series with the resistor" driven "by the presence or absence of these two signals."
• Address signal reaching the firing device's control input only when an enable permits — B's AND gate 48.
• The residual difference is purely topological: B gates with a logic gate at the input; '817 gates by a pass FET pulling the address into the first FET's gate node. This is a textbook substitution, and '817's own specification supplies the functional justification that makes it obvious — the gate capacitance of the (large) firing FET is charged by the address line and "acts as a storage device or memory device that retains an activated state" (FIGS. 9/11). Using a MOSFET gate node as a capacitive sample-and-hold is a conventional and well-understood circuit function; nothing in the references or in '817 suggests the substitution was unexpectedly difficult or produced an unpredictable result. This is the weakest of the grounds primarily because the exact pass-transistor-plus-capacitive-hold arrangement is not literally shown in A, B, or C, and the enable-hold/address-hold timing (T_AH, T_EH) that makes it work is not shown in the verified art. I would expect an obviousness challenge here to turn on whether the reference(s) or a fourth reference is found showing enable-gated address lines into a drive-FET gate.

Ground 4 — Claims 15, 17, 22–25 obvious as design choices / arithmetic

• Claim 15 (13 address + 16 drive + 2 enable, 16 simultaneous): arithmetic instantiation of Ground 1's formula. 13 × 2 = 26 per group; 16 groups = 416, and one-per-group concurrent firing = 16 simultaneous. Nothing in '817 attributes a criticality or unexpected result to the numbers 13/16/2 beyond the fact that they satisfy the formula and the bond-pad budget (the spec says only that "other arrangements are also contemplated" and that different numbers "may require different numbers" — a statement that the numbers are per se variable, i.e., design choice). In re Aller/In re Boesch line: a change in a parameter, or expression of an obvious result as a formula, is not inventive absent unexpected results.
• Claim 22/24 ("ratio … approximately 6.5 to 1") — 13 ÷ 2 = 6.5, a direct consequence of the numbers in claim 15. No evidence the 6.5:1 ratio is a result-effective variable producing a surprising property. The sibling '042 claims the same ratio, reinforcing that it is a carried-over arithmetic expression rather than a discovered optimum.
• Claims 23, 25 — (A × E × D) total generators. This is a counting identity, not an inventive step.
• Claims 7, 8 ("first switching device comprises a transistor"; "an NMOS transistor") — B teaches MOSFET switching devices; A teaches on-substrate integrated drive circuitry; the specification's own rationale for NMOS (fewer masking steps, no constant bias supply) is a known manufacturing/cost trade-off, and the claim recites the switch as an NMOS FET with no structural detail beyond the generic device.

Ground 5 — Claims 2, 3, 4, 11, 12, 13 obvious in view of A/B/C

• Claim 2/3 (resistive/thermal heating device in series with a series switch): A and B both.
• Claim 4 (second switching device in a current path between drive conductors): B's MOSFET 46 in series with resistor 44.
• Claim 11 (third switching device from the control terminal to one drive conductor, on a second enable): this is the "bleed/pull-down" device; C's auxiliary column drivers that "unaddressed heating element[s]" are driven by "auxiliary row and auxiliary column drivers which directly drive each unaddressed heating element with a specific voltage" is analogous, as is the standard practice of pull-down/bleed transistors on a drive-FET gate. This element rests on analogous art and on the PHOSITA's general knowledge; I would rate it moderate-confidence for obviousness absent a cleaner reference.
• Claim 12 (third/fourth switching devices giving two heaters gated by the same enable source, "configured to activate only one of the first and third … at the same time"): the one-at-a-time rule is A's explicit teaching; the complementary enable wiring is the predictable circuit realization of that rule.
• Claim 13 (plurality of first/second switching devices bound to different drive-current pairs and different address terminals): B's matrix.

5. Where the obviousness case is weak — and the best nonobviousness arguments

I should state these plainly, because a § 103 opinion that only argues for rejection is not useful.

1. The pass-gate topology and the capacitive-memory insight are genuinely elegant. '817's core insight — that the large drive FET's gate capacitance can replace a storage element so that the address and enable need not be asserted for the full drive pulse, shortening the time slot and raising print rate (FIGS. 9/11; "the gate capacitance … acts as a storage device or memory device") — is not found in A, B, or C as verified. A and B both require the address/enable to be present for the duration of the firing event (B's AND gate must hold both inputs; A's address enable must be held while primitive select is on). '817 reverses that timing constraint. If a claim were drafted to the timing/storage relationship (claim 10 arguably captures the structure but not the timing), the nonobviousness case would be materially stronger.
2. Image-independent, periodic address and enable signals. '817 makes address and enable purely periodic and image-independent, moving all image data onto the drive-current (P) lines. B's print enable signal is data-derived (register 40 receives the print command). This is a real architectural difference, though (as noted) it is not claimed in claims 1/14/16/18/19/20, which recite only "enable signals" / "select signal" generically — a drafting gap that makes the claims—not the invention—the weak point.
3. NMOS-only, no constant power supply. A significant commercial/technical advantage (device cost, masking steps, no bias circuit). But advantage-in-spec is not claim scope; and B already contemplated a constant-power interconnection, cutting toward obviousness of the power architecture.
4. Claim 26 is unanalyzed (source text truncated at "An inkjet printhea…"). I cannot assess it. Likewise, if claims 27–29 exist, they are unseen.

6. Secondary considerations

• Commercial significance / litigation: '817 was asserted by Hewlett‑Packard in ITC Inv. Nos. 337‑TA‑711 and 337‑TA‑723, N.D. Cal. Nos. 5:10‑cv‑02175, 3:10‑cv‑02175, 3:10‑cv‑00965 (HP v. Microjet Technology et al.), and went up on CAFC No. 12‑1225. Family-level litigation is evidence a patentee will press for nexus between the claimed subject matter and the commercial success — but litigation alone is not secondary-consideration evidence, and the numerical/formula claims (15, 16, 22–25) would likely fail the nexus requirement because the spec itself disclaims criticality of those numbers.
• No unexpected-results evidence identified in the fetched specification, and no data showing the 6.5:1 ratio is a result-effective variable.
• No evidence of copying, licensing, or long-felt-but-unmet need located in this review.

7. Bottom line

The decisive vulnerability of '817 is that US 5,541,629 (Saunders et al.) already discloses the same three-signal (power/address/enable) matrix and the same stated objective ("increase the number of nozzles without increasing interconnections"), and US 5,604,519 (Keefe et al.) already supplies the one-firing-element-per-shared-node rule and its crosstalk rationale. Combining the two to arrive at an enable-selected pair sharing address and drive lines is the predictable use of known elements for their known functions, and the numeric limitations (13/2/16; 6.5:1; A×E×D) are arithmetic consequences of that combination rather than inventive departures.

Confidence caveat: Because I could not retrieve '817's own examiner-cited reference list (the fetching gap flagged in §1), the grounds above are constructed, not the prosecution record. Before relying on this analysis for a validity challenge or an IDS, the actual front-page references and any EP counterpart search report (EP 1 330 358 B1, family member per the Google Patents "Also Published As" data) should be pulled, and claim 26 and any claims 27–29 must be read from the printed patent.