US 6310639

I'll search for this specific patent number and any related litigation.

Let me run a few more targeted searches to check for any 2026 litigation or administrative activity.

US Patent 6,310,639 — Analyst Summary

Source of record: Google Patents (USPTO full-text) page for US6310639B1, cross-checked against Justia Patents and FreePatentsOnline. All identifiers below are transcribed literally.

Bibliographic data

Family / continuity (per the patent's own "Description" and the Google Patents family table):

• Continuation-in-part of Ser. No. 08/738,516 (filed 1996-10-28, now US 6,113,221), itself a CIP of Ser. No. 08/597,746 (filed 1996-02-07, now US 6,000,787);
• And of Ser. No. 09/240,286 (filed 1999-01-29, now US 6,155,670), itself a CIP of Ser. No. 08/812,385 (filed 1997-03-05, now US 6,099,108).
• Foreign priority/parallel filings from 2000-04: EP 00303247 A → EP 1 048 465 B1, DE 60025890 T2, KR 10-2000-0021888 A → KR 100783976 B1, CN 001081020 A → CN 1170676 C. A later US continuation, Ser. No. 09/800,873 → US 6,540,325 B2.

Abstract (as published)

"An inkjet printing device is arranged to employ a first set of multiple nozzle drop generators activated by a first address signal and a second set of multiple nozzle drop generators activated by a second address signal. The multiple nozzles of each drop generator of the first set are arranged in a predetermined geometric pattern, each of which encompasses at least one nozzle of a drop generator of the second set. The ink ejectors of one drop generator of the first drop generator set are arranged in subgroups, one subgroup of which shares a switched power return with one subgroup of ink ejectors of one drop generator of the second drop generator set."

Plain-language overview of the independent claims

Claim 1 — Inkjet printing device (apparatus). Two "drop generators" (each = a group of ≥2 nozzles plus their heater-resistor ink ejectors that fire together on one command). The nozzles of the first form a geometric pattern; the nozzles of the second form a second pattern, and at least one nozzle of the second generator sits on or inside the perimeter of the first pattern (i.e., the nozzle groups interleave/overlap). On the drive side, a first switch feeds a primitive-signal input; second and third switches connect to a primitive return. One ejector of the first generator is coupled to switch 1 and switch 2; one ejector of the second generator is coupled to switch 1 and switch 3 — so the two generators share an input switch but have independently switchable returns.

Claim 4 — Inkjet printing device. Essentially the overlap-recitation variant of claim 1, explicitly requiring that the first "signal" comprise an address signal and an ink ejector primitive signal, with the second generator's nozzle(s) on or within the first pattern's perimeter.

Claim 5 — Inkjet printing device. A drive-architecture claim independent of nozzle overlap: a first drop generator is switchably coupled to a first ink-ejector primitive signal and, independently, to a first primitive-return conductor; a second drop generator is switchably coupled to the same primitive signal and, independently, to a second primitive-return conductor. The ejectors of the two generators are arranged in respective geometric patterns and disposed adjacent to one another.

Claim 13 — Method of depositing ink dots. (a) Simultaneously fire all ejectors of a first generator → a first plurality of dots in a first geometric pattern; (b) fire all of a second generator → a second plurality of dots in a second pattern, at least one of which lands on or within the first pattern's perimeter; (c) reposition the generators relative to the medium; (d) fire at least one but fewer than all ejectors of the first generator; (e) likewise for the second generator. This is the grayscale/dot-count modulation method.

Claim 14 — Method of depositing ink dots. Same "all, then fewer-than-all after repositioning" sequence without the geometric-pattern/perimeter limitations.

Claim 15 — Method of depositing ink dots. All-ejector firing of two generators with the perimeter-overlap limitation and repositioning, then firing fewer-than-all ejectors of the first generator only (no second-generator partial-fire step).

Claim 16 — Method (primitive-signal framing). The switching-based version of the deposition method: all ejectors of the first generator are switchably coupled to the primitive input and to the return to expel a first plurality of droplets; same for the second generator; reposition; then re-couple all ejectors to the input but only some of the ejectors to the return for each generator (independently controlled split returns).

Claim 18 — Method of manufacture. Building the device: arrange the simultaneously-energized nozzles/ejectors of generator 1 into a first geometric pattern and of generator 2 into a second pattern with at least one of generator 2's nozzles on or within generator 1's pattern; couple at least one ejector of each generator to a first switch; couple that first switch to a primitive-signal input; couple an ejector of generator 1 to a second switch and of generator 2 to a third switch; couple the second and third switches to a primitive-signal return.

Claims 21–22 are also method-of-manufacture claims (claim 21 begins identically to claim 18's preamble); the fetched record truncated their full text, so I will not characterize them beyond that.

Dependent claims add: four nozzles per generator in a parallelogram pattern (claims 2, 9, 20); "extended pixel" simultaneous ejection (claim 3); a third generator on a second primitive signal sharing the second return (claim 6); sets of generators (claim 7); nozzle-overlap-on-perimeter (claim 8); address-signal and return-activation-signal coupling (claims 10–11); coincident activation for extended-pixel deposition (claim 12); and third/drop-generator variants of the manufacturing method (claim 19).

CAFC 2026 docket check — negative result

I searched for a Federal Circuit 2026 docket entry naming 6,310,639 and found none. Publicly indexed results returned no CAFC appeal or § 337/ITC action for this patent. This is consistent with the record: the patent's stated legal status is Expired – Fee Related, with anticipated expiration 2016-02-07, nearly a decade before the date asked about. The only HP inkjet ITC enforcement matter surfaced in searching (In re Certain Inkjet Ink Cartridges with Printheads, USITC, instituted 2010, review of June 10, 2011 ID) asserted different patents — '598, '053, '347, '817, '279, and '917 — and not '639.

Uncertainty flags:

• I cannot fully exclude a non-indexed or unreported 2026 filing; my statement is limited to what public search surfaces.
• One aggregator page (patentleaderboard.com) lists this patent under a "Naoto Kawamura — 114 Patents at Canon" leaderboard. That attribution conflicts with the USPTO assignment record (Hewlett-Packard → Hewlett-Packard Development Co. L.P.) and with the patent's own text. I treat it as an unverified aggregator artifact and did not adopt it. The same page's "Approx Value" figure is an estimate, not an official valuation.
• Claim 21 and claim 22 text was truncated in the source record I fetched; treat my descriptions of those two claims as incomplete.

I'll search for litigation involving US patent 6310639.

The first hit refers to an unrelated patent (a tubular makeup "‘063" reissue, not US 6,310,639). Let me refine.

Initial searches surfaced only unrelated patents. Let me try more specific queries.

Search Results: Litigation Involving U.S. Patent No. 6,310,639

Bottom line: I found no litigation — district court, ITC, or PTAB — that specifically names U.S. Patent No. 6,310,639 ("Printer printhead," Kawamura & Weber, Hewlett-Packard) as an asserted, challenged, or otherwise-in-suit patent.

What I searched

I ran queries against general web sources indexing litigation databases (Unified Patents litigation pages, Justia/Federal Register ITC notices, Docket Alarm/CAFC materials, PACER-derived dockets) using the literal identifiers "6310639," "6,310,639," "US6310639B1," and combinations with "Hewlett-Packard," "inkjet," "infringement," and "lawsuit." Queries were run literally; I did not auto-correct any number.

What the results actually showed (all negative for '639)

No case found to report

Because no case naming the '639 patent was located, I cannot provide plaintiff, defendant, jurisdiction, case number, filing date, or status for any case — there is nothing to populate.

Important caveats

• Coverage limits. My search hit a tool step limit before I could exhaustively check paid docket services (Docket Navigator, LexisNexis IP Litigation, Westlaw Litigation Analytics) or the full PACER/Unified Portal case lists. I cannot certify a universal negative; I can only report that no publicly indexed litigation tied to the literal number 6,310,639 was found.
• Practical reason for scarcity. Per the source record, the '639 patent's priority date is 1996-02-07, and it reached its anticipated expiration on 2016-02-07, with legal status "Expired – Fee Related." Any infringement action would have had to be filed before that expiration, and the patent's owner of record (Hewlett-Packard Co. → Hewlett-Packard Development Company, L.P.) is an operating company rather than a litigious NPE, making a large public docket footprint unlikely. This is context, not evidence of a specific suit.
• Distinguish related family patents. The '639 patent is a continuation-in-part of U.S. 6,113,221 (itself a CIP of U.S. 6,000,787) and of U.S. 6,155,670 (CIP of U.S. 6,099,108). Litigation, if any, may have been brought on those family members rather than on '639 itself — but that would be a separate inquiry, and I found no case tying any of those numbers to the '639 patent either.
• Correction noted. The earlier search hit on a "‘063 patent" in the PMR/Frank's Casing appeals is a distinct patent and must not be conflated with U.S. 6,310,639; the alphanumeric similarity is coincidental.

Suggested next steps for a definitive check

1. Unified Patents Portal – https://portal.unifiedpatents.com/litigation/caselist (search by patent number; free).
2. USPTO Patent Center / Global Dossier for a full family cross-reference.
3. ITC EDIS (edis.usitc.gov) for any Section 337 complaint listing '639.
4. PTAB filings (IPR/PGR/CBM) via USPTO PTAB End-to-End, in case of a validity challenge that did not involve district court.
5. Paid docket services for a certified negative if this is for a formal opinion or clearance.

If you can confirm that the litigation you have in mind is against a specific defendant (e.g., a cartridge remanufacturer), give me the party name and I will re-run the search targeted to that docket.

I'll start by verifying the canonical "no proceedings" signal against the public record via web search.

Let me run additional targeted checks on PTAB dockets and any litigation that would have driven petitions.

Let me check whether the ITC respondents filed any PTAB/reexam challenges against the related HP printhead patents, and confirm the absence of proceedings on this patent's family.

Proceedings overview

Zero AIA trial proceedings are on file for US 6,310,639. The canonical structured block from the USPTO Open Data Portal reports no AIA trial proceedings, and my independent web checks surfaced none — the breakdown is therefore 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution-denied. The bottom line for a defendant is not the usual "the patent has been hardened by surviving IPRs" story and not "claims 1–5 are canceled." It is a more basic and more useful fact: no petitioner has ever taken this patent to the Board, and the patent itself is expired — Google Patents records an anticipated expiration of 2016-02-07 and a current legal status of "Expired - Fee Related" (US6310639B1). A demand letter citing US 6,310,639 today is citing a dead patent.

There is no proceeding to profile at claim-level granularity, so the per-proceeding template below collapses to a null entry plus the closest adjacent activity, which is worth knowing because it is what a careless analyst would mistake for PTAB history.

No proceeding to report

• Count: 0 IPRs, 0 PGRs, 0 CBMs, 0 derivation proceedings.
• Type: n/a
• Filed: n/a
• Status: n/a — ODP returns no AIA trials; corroborated by web search.
• Verbatim from structured data: "The USPTO ODP API returns no AIA trial proceedings for this patent as of the most recent ingest."
• Why the usual trial types could not have applied: this is a pre-AIA patent (priority 1996-02-07; application 09/300,785 filed 1999-04-27), so PGR was never available (PGR reaches only patents with an effective filing date on or after 2013-03-16). CBM was never available — the claims are directed to thermal inkjet drop generators and printhead architecture, not "a financial product or service," and the CBM program sunset on 2020-09-16. That leaves IPR as the only realistic AIA vehicle — and none was filed.

Closest adjacent activity (not PTAB, and not this patent)

• ITC Inv. No. 337-TA-723, Certain Inkjet Ink Cartridges With Printheads and Components Thereof, instituted 2010-06-18 against MicroJet, Asia Pacific Microsystems (APM), PTC, and the Mipo/SinoTime/Mextec entities. The asserted patents were US 6,234,598; 6,309,053; 6,398,347; 6,481,817; 6,402,279; and (withdrawn) 6,412,917 (75 FR 36442). US 6,310,639 was not asserted. ALJ Rogers found no invalidity as to the asserted patents (ID 2011-06-10); the Commission reviewed only the inducement finding, reversed it as to APM, and issued a general exclusion order (Commission opinion, Pub. 4373). APM's appeal was disposed of on waiver grounds at the Federal Circuit (Asia Pac. Microsystems v. ITC, No. 2012-1225).
• Relevance to '639: this is useful context, not a '639 proceeding. It shows HP was an active enforcer of a sibling printhead patent family around 2010–2011 and that the respondents mounted invalidity defenses (obviousness, enablement, written description) — yet '639 was left out of the complaint entirely, and no respondent petitioned the Board on it. If you are facing a '639 demand, that history tells you the patent was not the crown jewel in HP's own enforcement campaign.

Strategic summary

Claim status: wholly UNTESTED. No claim of US 6,310,639 has ever been construed by the PTAB, canceled, confirmed, or amended in any AIA proceeding. Claims 1–22 stand exactly as they issued on 2001-10-30 — not because they were defended successfully, but because they were never attacked. Because of that, there is no certificate of correction, no IPR certificate canceling any claim, and no FWD whose disposition I can quote. Anything asserting otherwise about this patent would be fabricated.

Estoppel landscape: essentially empty, but mooted by expiration. Section 315(e)(2) estoppel attaches only to a petitioner that reaches a Final Written Decision under § 318(a). Since no IPR produced an FWD here, no party — HP included — carries any § 315(e)(2) estoppel on this patent, and there is no § 325(e) PGR estoppel or AIA § 18 CBM estoppel either. In the abstract, that means every § 102/§ 103 ground based on patents and printed publications is still available to a defendant in district court or the ITC, with no PTAB-side washout. The practical point, however, is that the reason to care about prior art is prospective invalidity, and the reason to care about invalidity evaporates once the patent term ends.

Pattern signals: the absence is the signal. Same-petitioner serial filings — none. Patent-owner appellate aggression from PTAB loss — not applicable. Defensive aggregator (Unified Patents et al.) involvement — none found. On the record I could retrieve, HP never asserted '639 in its 2010–2011 inkjet-cartridge enforcement campaign at the ITC, no respondent counter-petitioned on it, and I found no district court case asserting it. Google Patents further records the patent as expired on 2016-02-07 with status "Expired - Fee Related." Note the patent also carries a live family — US 6,000,787, US 6,113,221, US 6,099,108, US 6,155,670 and US 6,540,325 are all named as priority parents/children — and my searches did not turn up IPRs on those either, though I cannot rule that out exhaustively.

Recommended next steps

1. Lead with the expiration, not the PTAB record. Before spending on any validity analysis, confirm the term and any PTA on the face of the patent and via USPTO Patent Center. The ODP/Google data show priority 1996-02-07, grant 2001-10-30, anticipated expiration 2016-02-07, legal status "Expired - Fee Related." If that holds, post-2016-02-07 activity cannot infringe at all, and any past-damages theory runs into the § 286 six-year lookback — which, given expiration eleven years ago, likely leaves nothing recoverable.
2. Do not let opposing counsel conflate this patent with its siblings. The ITC win in 337-TA-723 involved a different set of HP patents (the '598, '053, '347, '817, '279, and '917). If a demand letter leans on that litigation or its general exclusion order, ask them to identify the specific claim of '639 they assert and the accused product's post-2016-02-07 acts. The 337-TA-723 exclusion order does not cover '639.
3. If you are nonetheless challenged on validity (e.g., a residual past-damages claim), treat the field as wide open: no § 315(e)(2) estoppel binds anyone, so the full universe of patents and printed publications is available under § 102/§ 103, including the art already cited against the family. Note that the '639 specification itself concedes a crowded art — it cites U.S. Pat. Nos. 4,967,203, 4,999,646, 5,583,550 and 5,541,629 as known techniques, and lists US 4,746,935, US 5,648,805, US 5,648,806 and US 5,984,464 among cited references (Justia claim/description text, Google Patents).
4. State the PTAB conclusion plainly in any internal memo: there is no PTAB proceeding to cite, no FWD to link to, and no institution decision to quibble with. If a third party hands you an IPR number purporting to be against US 6,310,639, treat it as a typo or a misidentification of a sibling patent until the number checks out on PTAB E2E.

Caveat on scope: I could not query PTAB E2E or the ODP API directly from this session; the "no proceedings" conclusion rests on the canonical structured block supplied in the prompt, corroborated by targeted web searches that returned no PTAB docket entries for US 6,310,639 or any of its family members. If you need a belt-and-suspenders confirmation, run the patent number through PTAB E2E (Patent Trial and Appeal Board End-to-End) and the CourtListener docket search before finalizing.

I'll research the assignment record for US 6,310,639. Let me start with parallel searches.

Let me dig deeper on litigation history and any post-2005 transfers.

Ownership Chain Reconstruction — US 6,310,639 B1 ("Printer printhead")

Retrieval caveat up front. Google Patents' "Legal Events" table for this patent surfaces exactly two reassignment events (both described below). No reel/frame numbers are exposed by that source, and I was unable to complete a live query of assignmentcenter.uspto.gov in this session (JavaScript application; the legacy mirror at assignment.uspto.gov/patent/index.html returns the same underlying records). I therefore report the events with dates and parties as recorded, and explicitly mark reel/frame as not retrieved rather than guessing. Verify at:

• https://assignmentcenter.uspto.gov/ (search patent number 6310639)
• https://assignment.uspto.gov/patent/index.html

Inventors

Pattern notes / caveats:

• Both inventors reside in Corvallis, Oregon, the location of HP's thermal-inkjet printhead design and manufacturing operation in that era; the companion applications in this family (US 6,000,787; 6,113,221; 6,099,108; 6,155,670; 6,540,325) name the same two inventors, so this is a focused, single-team family rather than a distributed portfolio.
• Departure within 12 months: no data. I found nothing in the sources I could reach indicating either inventor left HP near filing. Not inferred either way.
• Data-quality flag: a third-party aggregator (patentleaderboard.com) groups a "Naoto Kawamura" under Canon with 114 patents. That grouping conflicts with the face of this patent, which names Hewlett-Packard Company, Palo Alto, CA as assignee, and with the recorded HP assignment. I treat the Canon grouping as an artifact of that site's ownership-scraping and do not adopt it. The same site's "$16,862,000" figure for this patent is an algorithmic estimate, not evidence of value.

Original assignee

Hewlett-Packard Company, Palo Alto, California — a Delaware corporation.

• Primary line of business: at issue date (2001-10-30) the largest printer manufacturer in the world; thermal inkjet printing was a core business line.
• Did it ship product embodying the claims? Yes, on the balance of the evidence. The patent is directed to a multi-nozzle drop generator in which the nozzle pattern of one drop generator overlaps the pattern of a neighboring generator (FIG. 4C; claim 1's "at least one nozzle associated with said second drop generator is disposed on or within the perimeter of said first geometric pattern"), with split primitive-return switching to fire fewer than all nozzles of a generator. That is HP thermal-inkjet printhead architecture of the late-1990s/early-2000s (DeskJet / OfficeJet / DesignJet class cartridges). I could not pin the specific SKU, and I state that as an inference from the technical disclosure, not a documented product mapping.
• Current status: the named assignee entity is no longer an independent operating company. Hewlett-Packard Company separated on 2015-11-01 into HP Inc. (personal systems and printing) and Hewlett Packard Enterprise Company. The printing business and this technology line went with HP Inc. I could not confirm a US recording that specifically reflects this separation for US 6,310,639; the only US transfer recorded against it in the sources I reached is the HP → HP Development Company, L.P. entry below. Not in bankruptcy; no Chapter 7/11 proceeding found.

Assignment timeline

Two recorded events. Both are confirmed by Google Patents' legal-events table. Reel/frame not retrieved for either (see caveat above) — flagged as such, not fabricated.

• 1999-06-15 (recorded) — Reel not retrieved

  • Conveyance: Assignment of assignors' interest (inventor → company)
  • Assignor: Kawamura, Naoto A.; Weber, Timothy L.
  • Assignee: Hewlett-Packard Company (Palo Alto, CA)
  • Correspondent: not retrieved. No NPE-associated outside counsel is indicated anywhere in this chain; HP matters of this vintage were customarily recorded through HP's internal Intellectual Property Administration (3404 E. Harmony Road, MS 35, Fort Collins, CO 80528 — the address that appears as HP's correspondence address of record on sibling HP patents). I flag this as expected-not-verified for this specific reel.
  • Context: Standard employee invention assignment; recorded ~7 weeks after the 1999-04-27 filing of application 09/300,785. No third party involved.

• 2005-01-11 (recorded) — Reel not retrieved

  • Conveyance: Assignment of assignors' interest
  • Assignor: Hewlett-Packard Company
  • Assignee: Hewlett-Packard Development Company, L.P. (HPDC; Houston, TX limited partnership)
  • Correspondent: not retrieved; filed through HP's own IP administration rather than an outside law firm, consistent with the pattern of HP's mass intra-group recordings.
  • Context: Internal corporate reorganization, not a sale. This is HP's company-wide transfer of its patent estate into its IP-holding limited partnership. Corroboration from the parallel EP/GB family records: the UK register shows HPDC "registered as Applicant/Proprietor in place of HEWLETT-PACKARD COMPANY … by virtue of deed of assignment dated 31.01.2003." That is the same corporate deed; the US-side recordation appears as 2005-01-11. HP ran comparable bulk recordings into HPDC in this window (and, for the enterprise portfolio, later out of HPDC to Hewlett Packard Enterprise Development LP — e.g. reel 037079/0001, executed October 2015, per a published assignment exhibit).

Nothing further is recorded against US 6,310,639. Google Patents' legal status is "Expired – Fee Related" with an anticipated expiration of 2016-02-07 — i.e., the patent lapsed for non-payment of maintenance fees and never reached full term.

Related family members (US 6,000,787, 6,113,221, 6,099,108, 6,155,670, 6,540,325) share this same inventor team and the same HP → HPDC ownership posture; none of them shows a third-party assignment in the sources I reached.

Timeline diagram

timeline
    title Ownership of US 6310639
    1996 : Priority date from parent case
    1999 : Filed as continuation in part
         : Inventors assign rights to Hewlett-Packard
    2001 : Patent issued Oct 30
    2005 : Transfer recorded to HP Development Co LP
    2015 : Hewlett-Packard separates into HP Inc and HPE
    2016 : Patent lapses for unpaid maintenance fees

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The only transferee in the chain is Hewlett-Packard Development Company, L.P., a Texas limited partnership that is HP's own IP-holding subsidiary, not a licensing-only shell. No "IP / Patents / Licensing / Ventures" third-party entity appears; no registered-agent-service address; no single-member Delaware/Texas LLC. Recorded 2005-01-11 (reel not retrieved).
2. Known asserter in the chain — not present. No assignee or assignor matches Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, or any Spangenberg entity. Chain is Hewlett-Packard Company → HP Development Company, L.P. throughout.
3. Repeat correspondent across the chain — unclear. I could not retrieve the correspondent of record for either recording, so I cannot test for recurrence. What I can say: nothing in the visible record points to an outside NPE-side filing attorney; the recordings are consistent with HP's internal IP administration (Fort Collins, CO) handling both events itself. Treat as not evidenced, not as exculpatory.
4. Cascading transfers — not present. One intra-group transfer in 24 years of recorded title; no chained LLC sequence, no shared-principal pattern to assess.
5. Pre-litigation transfer — not present. No infringement suit naming US 6,310,639 was found. The most recent transfer of record (2005-01-11) predates the patent's 2016 lapse by more than a decade; there is no transfer within any 6-month window before any suit. (Note: an unrelated Bard "’639 patent" appears in a MedComp v. Bard antitrust/Walker Process decision — that is a different patent, not US 6,310,639, and I did not treat it as evidence.)
6. Bankruptcy fire-sale — not present. No Chapter 7/11 filing by HP or HPDC touching this patent; no sale in bankruptcy proceedings.
7. Privateering — not present. No transfer to an NPE asserted for HP's benefit; no SEC-filing or Patent Progress/EFF coverage linking this patent to a privateering arrangement.
8. Defensive aggregator — not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at HPDC, an operating-company-affiliated IP holder, and the patent then lapsed for failure to pay maintenance fees — an inverse of the aggregator signal, and itself evidence that no one perceived assertion value in it.

Score: 0 strong signals, 0 weak signals, 1 unclear (correspondent data unavailable).

Verdict

Insufficient data — with the qualification that the limited record that does exist affirmatively shows an operating-company-only chain.

Justification: the recorded chain is two links and no more — the 1999-06-15 inventor-to-Hewlett-Packard assignment (Kawamura and Weber) and the 2005-01-11 HP-to-Hewlett-Packard Development Company, L.P. intra-group recording, both consistent with HP's company-wide reorganization (corroborated by the UK register's "deed of assignment dated 31.01.2003"). There is no third-party transferee, no asserter name anywhere in the chain, no litigation naming this patent, and the patent lapsed for unpaid maintenance fees on 2016-02-07 ("Expired – Fee Related"). No NPE signal is present; the only gap is that I could not retrieve reel/frame or correspondent data, which is why the rubric's "insufficient data" bucket is the honest choice rather than an affirmative NPE or defensive-aggregator call.

Verify the underlying records directly: https://assignmentcenter.uspto.gov/ (patent number 6310639) and the legacy mirror https://assignment.uspto.gov/patent/index.html.

I'll search for the patent's citation record to identify the prior art references cited on its face.

Let me retrieve the full citation record directly from the patent page.

Let me retrieve the full front-page citation list from the patent's PDF and FPO record.

Prior Art Analysis — U.S. Patent No. 6,310,639 ("Printer printhead")

Scope and source note. I searched for the literal identifiers "6310639," "6,310,639," "US6310639B1" and pulled the citation record from the Google Patents entry for US6310639B1 (which mirrors the USPTO face-of-patent "References Cited" plus references cited in the specification). I did not retrieve results for any other number, and no similar-number patents are included below.

Important completeness caveat — read first. The face-of-patent citation table for '639 appears to list 14 patent documents. My searches surfaced nine (9) patent-citation rows in full, plus the applicant-cited references named in the specification's Description. I could not retrieve the complete 14-row table (the table was truncated at the row for EP 0 913 257 A2 in every source I reached). I therefore flag below which entries are confirmed from the citation table, which are confirmed from the specification text, and which could not be confirmed. I did not fabricate any citation, date, or description.

A. Confirmed patent citations (from the Google Patents "Patent Citations" table for US6310639B1)

Dates below follow that table's columns (priority/inventory date, then publication/publication-or-grant date). Where the same family has multiple members, they are listed separately because they appear separately on the face of the patent.

B. Applicant-cited references named in the '639 specification (Description)

The specification expressly cites these; they are the references most directly tied to the two inventive themes of '639 (dot-placement/interleaving, and IDH multiplexing). Dates are from the granted patents / citation records I could verify; where I could not verify a date in this session I say so.

Unresolved gap: the 14-row face-of-patent table is 9 rows confirmed above; entries 10–13 are confirmed as specification-cited references, but whether all four appear in the face-of-patent table (and whether one further citation exists) could not be confirmed because the table truncated. Treat rows 10–14 as not fully verified as to their exact placement on the face of '639.

C. § 102 anticipation analysis (claim-by-claim)

Framework. '639 issued from an application filed 1999-04-27, claiming earliest priority 1996-02-07. It is therefore governed by pre-AIA 35 U.S.C. § 102. Anticipation under § 102 requires every element of a claim disclosed in a single reference, arranged as in the claim. Note the CIP structure: method/apparatus claims supported only by new matter may be entitled to no earlier than 1999-04-27, while claims supported by the '746/'516 and '385/'286 parents reach back to 1996-02-07 at the earliest. The § 102(b) one-year bar and § 102(e) prior-art date therefore shift depending on which claims are analyzed — a point worth resolving against the priority proofs before any formal opinion.

Claims concerned with the drive/switching architecture — claims 1, 5, 6, 10, 11, 16, 18–22

• US 5,541,629 (Saunders/HP) — most relevant to claims 5, 10, 11, 16, 18.
  Saunders discloses the full IDH vocabulary the claims recite: primitives, a primitive select input, a primitive return/common, shared address lines, and a switch (MOSFET) in series with each heater resistor. This maps to claim 5's "switchably coupled to a first ink ejector primitive signal" and "switchably coupled to a primitive return conductor," and to claim 10's address signal and claim 11's return activation. It is the strongest single-reference candidate for the primitive-signal/return coupling limitations of claims 5, 10, 11, 16, and 18.

  • Gap: Saunders does not, on the record I retrieved, disclose the '639-specific feature of two independent switched primitive returns per cell with a shared input switch (claim 1's "second and third switches coupled to a primitive signal return"; claim 5's first vs. second return conductor independence), nor the overlapping nozzle-pattern-on-perimeter feature. So it cannot alone anticipate claim 1 or claim 4 as a whole.

• US 5,648,805 (HP, "Inkjet printhead architecture for high speed and high resolution printing") — also relevant to claims 5, 6, 10, 11, 16, 18.
  Filed 1992-04-02 and granted 1997-07-15, this is § 102(e) prior art against any claim not entitled to the 1996-02-07 date (and § 102(a)/(b) art against the 1996 date if the priority proofs are limited). It is directed squarely at printhead drive architecture (primitives, select/return, addressing, switches) and is a natural single-reference candidate for the architecture limitations of claims 5, 6, 10, 11, 16, and the manufacture-coupling steps of claim 18.

Claims concerned with nozzle-pattern overlap — claims 1, 4, 8, 9, 12, 13, 15, 17, 20

• None of the nine confirmed citations is a clean single-reference anticipator of the "at least one nozzle of the second drop generator disposed on or within the perimeter of the first geometric pattern" limitation. The two HP substrate/nozzle-array patents (US 5,648,806 and its continuation-type sibling US 5,984,464) address a stable substrate structure and wide-swath nozzle array — relevant context for arranging many nozzles densely, but I did not retrieve disclosure of the specific interleaved/overlapping drop-generator perimeters claimed in '639. They are best characterized as § 103 combination art, not § 102 anticipators, for claims 1, 4, 8, 9, 13, 15, 17, 20.
• EP 0 897 804 A2 (Xerox, "Liquid ink printhead," pub. 1999-02-24) is a nozzle/channel printhead reference whose publication (1999-02-24) post-dates the 1996-02-07 priority; against claims entitled to that date it is not prior art at all, and against claims entitled only to 1999-04-27 it is prior art only if the application's effective § 102(a) date is on/after its publication — a date question that must be resolved per claim.

Method-of-depositing claims (dot-count modulation) — claims 13, 14, 15, 16, 17

• US 4,746,935 (HP, "Multitone ink jet printer and method of operation") is the most on-point confirmed citation for the "fire multiple drops / modulate number of dots per pixel" concept underlying the method claims. The '639 specification itself frames its contribution against exactly this multitone/multi-drop art (it cites the Doan, Trask and Hickman patents in the same paragraph). '935 is a plausible single-reference candidate for claim 14 (the broadest, purely "all-then-fewer-than-all after repositioning" method, with no perimeter limitation), and it is strong § 103 art against claims 13, 15, 16, 17 — but because '639's method claims add the "at least one but fewer than all ejectors after repositioning" and/or perimeter-overlap steps, whether '935 anticipates turns on whether it discloses selectively firing a subset of a multi-nozzle generator's ejectors on a later pass, which I could not confirm from the retrieved record.

Remaining confirmed citations (secondary relevance)

• EP 0 440 490 B1 (Canon, "Recording method and apparatus") — recording-method art; potentially § 103 material for the method claims 13–15, weak as a § 102 anticipator of the specific dot-count/perimeter steps.
• US 6,053,599 A (Canon, "Liquid jet printing head …") — printhead-structure art; § 103 material for claims 5, 7, 18–22.
• JP H06-079873 A (Fuji Xerox, "Ink-jet recording device," pub. 1994-03-22) — device-level art; § 103 material, not a clean § 102 anticipator of the claimed switch/return or perimeter features.
• JPH0679873A pre-dates both the 1996-02-07 priority and the 1999 filing, so it qualifies as § 102(b) art for claims reaching the earlier date.
• EP 0 913 257 A2 (HP) — title truncated and subject matter unconfirmed; I cannot responsibly assign it to a claim without the full text.

D. Ranking — most to least relevant prior art for '639

1. US 5,648,805 A (HP; 1992 filing / 1997 grant) — printhead drive architecture; best single-reference art for claims 5, 6, 10, 11, 16, 18.
2. US 5,541,629 A (Saunders/HP; 1996 grant) — IDH multiplexing; strongest art for the primitive-signal/return coupling of claims 5, 10, 11, 16, 18 (explicitly relied on by '639 itself).
3. US 4,746,935 A (HP; 1988 grant) — multitone/dot-count modulation; best art for method claims 13–17 (esp. claim 14).
4. US 4,967,203; US 4,999,646; US 5,583,550 (Doan; Trask; Hickman — HP) — interlace/dot-placement/dot-uniformity; § 102/§ 103 art for claims 13, 15, 17 and the overlap concepts.
5. US 5,648,806 A / US 5,984,464 A (HP) — substrate/nozzle-array structure; § 103 art for claims 1, 4, 8, 9, 20.
6. US 6,053,599 A (Canon), EP 0 897 804 A2 (Xerox), EP 0 440 490 B1 (Canon), JP H06-079873 A (Fuji Xerox) — secondary device/structure art.
7. EP 0 913 257 A2 (HP) — unresolved (incomplete record).

E. Explicit uncertainty flags

1. Incomplete citation set. Only 9 of the ~14 face-of-patent citations were retrievable; five rows (likely including the Doan/Trask/Hickman/Saunders description-cited patents and possibly one more) could not be confirmed as literal face-of-patent entries. I did not invent numbers to fill the gap.
2. Descriptions vs. disclosures. For references I could not open in full, the "brief description" is drawn from the title and citation context, not from a full-text elemental comparison. My § 102 statements are therefore preliminary, expressed as "potentially" and "candidate," not as established anticipations.
3. Anticipation vs. obviousness. Many of the true § 103 combinations here would require pairing a drive-architecture reference (Saunders / '805) with a dot-placement reference (Doan / Hickman / '935). Only US 4,746,935 and US 5,541,629 approach single-reference § 102 status for the broadest independent claims (14 and 5/16 respectively), and even those have gaps noted above.
4. Priority-date sensitivity. Because '639 is a CIP of four applications, the prior-art date for any given claim must be established per claim; my analysis assumes the 1996-02-07 earliest priority unless otherwise noted.
5. No litigation overlay. Consistent with the earlier litigation section of this analysis (which found no case naming '639), the above is a prosecution-history/face-of-patent prior-art review only — there is no litigation record to corroborate which references the patentee or a court treated as material.

If you can supply the full 14-row citation table (e.g., from the granted-patent PDF front page) or confirm the target claim(s)/priority date for a formal opinion, I can complete the elemental § 102/§ 103 mapping against the missing rows and tighten the anticipation calls.

I'll gather the prior-art record for this patent before running the § 103 analysis.

Let me pull the complete front-page reference list and the key primary references.

Now let me examine the key cited references in detail.

Obviousness Analysis — U.S. Patent No. 6,310,639 under 35 U.S.C. § 103

This section builds on, and does not repeat, the earlier "Patent summary" and "Litigation summary." Two findings from those sections materially frame the analysis below: (1) the '639 patent's legal status is "Expired – Fee Related," anticipated expiration 2016-02-07, so there is no live validity controversy to defend against; and (2) no litigation, ITC action, or PTAB proceeding naming '639 was located, meaning the following is a paper obviousness assessment grounded in the prosecution record, not a reconstruction of an actual validity fight.

1. Legal framework and the threshold question: what is the effective filing date?

Before any § 103 comparison can be made, the governing date for each claim must be fixed, because it determines the prior-art window.

Why this matters. The record shows the application was filed 1999-04-27 as a continuation-in-part, and the "Patent Citations (14)" list retrieved from the Google Patents record includes references whose dates sit between the 1996 priority date and the 1999 filing date — notably US 5,648,805 and US 5,648,806 (both issued 1997-07-15, from a 1992-04-02 priority) and the 1997-08-15/1997-10-30-priority items. The examiner's reliance on them is only coherent if the switched-return claims carry the 1999-04-27 date. Practitioners assessing this patent should assume the two-date split above unless the intervening files of the parent applications show otherwise. I have not reviewed those files; this is an inference from the citation dates and the CIP structure, and should be verified against the applications' respective as-filed disclosures.

2. Person having ordinary skill in the art (POSITA)

A POSITA here would hold a bachelor's degree in electrical or mechanical engineering (or equivalent) plus 2–4 years of experience designing thermal inkjet printheads, including thin-film heater-resistor arrays, orifice/barrier layer fabrication, and integrated drive head (IDH) multiplexing. Such a person would be familiar with:

• the Hewlett-Packard Journal printhead issues and Lloyd & Taub, Output Hardcopy Devices, ch. 13 — both cited in the '639 specification itself;
• IDH matrix drive conventions (primitive select / primitive return / address lines), which the '639 specification expressly admits are "conventionally used" and cites to U.S. Pat. No. 5,541,629; and
• the grey-scale / multi-drop-per-pixel literature the specification itself lists (US 4,967,203 Doan; US 4,999,646 Trask; US 5,583,550 Hickman).

That last point is important and deserves emphasis: the '639 Background of the Invention contains applicant-admissions that the entire matrix drive architecture and the multi-drop/tonal-printing objective are prior art. Under § 103 these admissions narrow the space in which non-obviousness could reside to (a) the nested nozzle-pattern geometry and (b) the split/switched primitive return sharing between drop-generator subgroups.

3. The prior art of record

From the Google Patents front-page citation list (https://patents.google.com/patent/US6922203B2/en#6, which reproduces the '639 citation table), the "Patent Citations (14)" entries I was able to recover are:

Caveat on completeness. The record shows 14 citations; I recovered nine before hitting retrieval limits. The missing five are most likely additional US/JP patent documents plus the non-patent literature the specification already names. Any of the two remaining independent claims' elements should be re-checked against the full list before this analysis is relied upon formally.

Expanded content of the two most important references (retrieved in full or near-full):

• US 5,541,629 (Saunders et al., HP) / EP 0 592 221 B1 — "Printhead with reduced interconnections to a printer." This is the reference the '639 specification itself cites. It discloses: drivers split into "primitives," each with its own primitive select (power) and primitive return (common); address lines shared among all primitives, "only one of which is preferably active at a time"; "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"; an XY matrix "where X is the number of primitives (rows) and Y is the number of drivers per primitive (columns)"; and — critically — alternatives including "energy control" (a circuit monitoring the voltage across each heater resistor and modulating the gate drive) and "gray scale modulation" via a multiplexer (60) feeding the input of each switching device with selectable analog levels GL1…GLn. It further states the object of "minimizing the number of interconnections required per driver" and reducing cost of switching supplies.
  • Sources: https://patents.google.com/patent/US5541629; http://www.everypatent.com/comp/pat5541629.html; https://patents.google.com/patent/EP0592221
• US 5,648,805 (HP) — "Inkjet printhead architecture for high speed and high resolution printing." Abstract/description retrieved: "The substrate contains two spaced apart arrays of ink ejection elements, and each orifice in the nozzle member is associated with a firing chamber and ink ejection element"; high-speed printing is achieved by "offsetting neighboring ink ejection elements from each other in each primitive grouping in the linear array"; 600 dpi "by densely positioning the ink ejection elements in each linear array"; and the driver circuitry comprises "an array of fourteen primitives, fourteen primitive commons, and twenty-two address select lines," with the express rule that "only one ink ejection element at a time in each primitive grouping" is fired "thereby minimizing undesirable interference such as fluidic crosstalk between closely adjacent ink firing chambers."
  • Source: https://patents.google.com/patent/US5648805A/en; abstract via https://pubchem.ncbi.nlm.nih.gov/patent/US-5648805-A

4. Element-by-element application to the independent claims

4.1 Claim 1 (device: nested nozzle patterns + shared input switch, split returns)

Combination 1 (primary): US 5,541,629 + US 5,648,805 (+ US 4,746,935).

Motivation to combine. Both references are Hewlett-Packard printhead patents addressing the same recognized problem — increasing nozzle count without a proportionate increase in printer-to-printhead interconnections — and both use the same architectural vocabulary (primitives, primitive select, primitive return, address lines). Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), where a technique has been used to improve one device and a POSITA would recognize that it would improve similar devices in the same way, the combination is obvious. The specific motivation for splitting the return into separately switched legs is supplied directly by the art: US 5,541,629 already teaches "gray scale modulation" and reduced interconnections, and US 5,648,805 already teaches multiple primitive-common conductors with the firing rule "only one … at a time in each primitive grouping" to control fluidic crosstalk and energy uniformity. A POSITA seeking to (i) fire a subgroup of a multi-nozzle drop generator (fewer than all) and (ii) keep the shared input switch count low would naturally place the selectivity on the return side rather than duplicate primitive selects — precisely the arrangement claim 1 recites. The result is a predictable rearrangement of known circuit elements performing their known functions (KSR; MPEP 2143.01).

4.2 Claim 4 (device: same overlap, with the "first signal" being an address signal and a primitive signal)

Substantively claim 1 minus the switch sub-architecture, plus an explicit two-signal enable. US 5,541,629 discloses exactly this two-signal enable: "where a primitive select interconnection and a control interconnection for a heater resistor are both active concurrently, that resistor is energized," and only one address line active at a time. US 5,648,805 confirms ("firing a particular resistor requires applying a control voltage at its 'Address Select' terminal and an electrical power source at its 'Primitive Select' terminal"). Combination 1 (minus the switch limitation) renders claim 4 obvious with high confidence.

4.3 Claim 5 (device: shared primitive signal, independently switchable first vs. second primitive-return conductors)

This is the broadest and most exposed independent claim, because it omits the nozzle-overlap limitation entirely and recites only (a) two drop generators, (b) each switchably coupled to the same primitive signal and to different primitive returns, and (c) their ejectors arranged in respective geometric patterns and "disposed adjacent" one another.

Combination 2 (for claims 5, 6, 7, 12, 18, 19): US 5,541,629 + US 5,648,805, optionally further with EP 0 913 257 (HP high-frequency ejection/refill, which is directed to the same problem of coordinating multiple ejection elements in a shared fluidic environment). The claim's "independently" language is met by merely providing a second switch in the return leg — a duplication of a known element (the MOSFET switch of '629) in a known circuit position (the return leg), with a predictable result (independent energization of subgroups). That is the archetype of an obvious design choice under KSR.

4.4 Claim 13 / 14 / 15 (methods: fire all, reposition, then fire fewer than all)

The specification itself concedes the building blocks: multiple drops per pixel "resulting in more 'levels' per color and reducing quantization noise," citing US 4,967,203, US 4,999,646, and US 5,583,550; and the background discusses interlacing/banding and "extended pixel" spacing.

• US 4,746,935 — "Multitone ink jet printer and method of operation" — is the of-record reference aimed at tonal (variable-dot-count) printing, the direct antecedent of the "activate at least one but fewer than all" step.
• The "repositioning" step is nothing more than the carriage scan along axis "X" and the media advance along "Y" that the '639 specification describes as conventional for any reciprocating inkjet printer (FIG. 1).
• The "geometric pattern" / "perimeter" limitations of claims 13 and 15 are met, if at all, by the same offset/interleaved-array teaching of US 5,648,805 discussed in § 4.1.

Combination 3: US 4,746,935 + US 5,648,805 + US 5,541,629. Motivation: the art recognizes a twofold desideratum — increase tonal levels (multitone/'935) while managing the electrical interconnection and crosstalk costs of a dense, high-frequency array ('805, '629). Selecting the number of ejectors fired per pixel by energizing a switched subgroup is the natural union of those teachings, and it is achieved without additional primitive selects.

4.5 Claim 16 (method in primitive-signal terms, with re-coupling of only some ejectors to the return)

Functionally identical to claims 13–15 phrased in circuit language; the added matter is "switchably coupling all of the ink ejectors … to the input … and at least one but fewer than all … to the return." That is claim 1's switch topology applied to the claim-13 method. Same combination (1 + 3) applies, plus the motivation that return-side selectivity avoids duplicating high-current primitive-select traces and switching supplies — a drawback US 5,541,629 identifies by name ("The primitive select interconnection for each primitive must be driven by a switching power supply that can rapidly switch between on and off states. Such supplies are more expensive and more prone to failure than constant, i.e., static, power supplies").

4.6 Claim 18 (method of manufacture)

Claim 18 recites, as manufacturing steps, exactly the structural couplings of claim 1. Method-of-making claims that recite only the steps of assembling the known structural elements of a device taught by the prior art are obvious for the same reasons as the apparatus claim. Combination 1 disposes of claim 18.

5. Dependent claims

6. Where the § 103 case is weakest — and the counterarguments a patent owner would raise

A. The "on or within the perimeter" limitation is the least well-supported element. I was unable to retrieve the full text of EP 0 897 804 A2 (Xerox), US 6,053,599 (Canon), JP H06-79873, US 5,984,464, US 5,648,806, EP 0 440 490, and US 4,746,935 before exhausting my retrieval budget. My mapping of "at least one nozzle of the second generator disposed on or within the perimeter of the first geometric pattern" rests on the offset/staggered neighbouring ejection elements language in US 5,648,805 and on the titles/abstracts of the others. That is a reasonable but unverified mapping. A rigorous invalidity contention would need those documents read in full. I flag this as the single largest evidentiary gap.

B. Teaching away — US 5,648,805's "one at a time" rule. US 5,648,805 teaches that high frequency with low crosstalk is achieved by "firing only one ink ejection element at a time in each primitive grouping." A patent owner could argue this affirmatively directs away from simultaneously energizing the ≥2 nozzles of a drop generator as claim 1 requires. Two rebuttals are available:

1. The reference's concern is fluidic crosstalk between adjacent firing chambers, not simultaneity per se; and the '639 specification solves precisely that problem structurally — its per-firing-chamber dedicated vias 321, 323 "encompassed by the firing chamber's lower periphery 319, so that … any pressure generated within the firing chamber will not generate ink flow to other chambers" — so a POSITA following '805's teaching would be motivated to adopt a multi-via/multi-chamber geometry enabling simultaneous multi-nozzle firing.
2. A teaching away must be such that a POSITA would be discouraged from following the path, not merely that a reference discloses an alternative (In re Gurley). Here the art's stated goal — more nozzles, more tonal levels, fewer interconnections — pushes toward the claimed arrangement.

C. Secondary considerations. Under In re Cree / WBIP, objective indicia can rebut a prima facie case. On the record before me there are none to weigh: the patent has been expired since 2016-02-07, and the earlier search found no litigation, ITC § 337 action, or PTAB challenge naming '639. There is therefore no evidence of commercial success, copying, industry praise, or long-felt-but-unmet need tied to these claims. Absent such evidence, the prima facie case stands unrebutted. (I have not searched for the HP commercial products — e.g., the DeskJet 820C-era tri-color printheads discussed in the June 1997 Hewlett-Packard Journal — for a nexus to these specific claims; that would be the obvious place for a patent owner to look.)

D. § 103(c) traps for the family references. Several of the most on-point documents are Hewlett-Packard's own — US 5,648,805, US 5,648,806, US 5,984,464, US 5,541,629, EP 0 913 257, and the parent applications US 6,000,787 / US 6,113,221 / US 6,155,670 / US 6,099,108. Under pre-AIA 35 U.S.C. § 103(c) (the statute governing this 1999-filed, pre-AIA application), commonly-owned art that qualifies only under § 102(e)/(f)/(g) is disqualified as prior art for § 103. Because the '639 patent issued from a Hewlett-Packard application and all of these were HP-owned at the relevant time, a contention built on the HP references alone may be procedurally vulnerable. US 4,746,935 (also HP, 1988) is not at risk — it is printed publication art under § 102(b). Practical consequence: a robust challenge should anchor on the non-HP members of the of-record art — JP H06-79873 (Fuji Xerox), EP 0 897 804 A2 (Xerox), US 6,053,599 (Canon), and EP 0 440 490 B1 (Canon) — combined with US 4,746,935.

7. Bottom line

The core § 103 theory is this: the '639 patent sits at the intersection of two well-developed, same-field, same-problem HP practices — IDH matrix multiplexing with primitives, primitive selects, and primitive commons (US 5,541,629, cited by the applicant in the specification as conventional), and dense, offset, high-resolution multi-element arrays fired under a one-per-primitive discipline (US 5,648,805) — plus the long-standing tonality objective of multi-drop printing (US 4,746,935, and the Doan/Trask/Hickman patents the applicant itself lists). Adding a second MOSFET in the return leg to independently select a subgroup of a multi-nozzle drop generator is the predictable, mechanism-preserving combination of those teachings, undertaken with a recognized motivation (fewer switching power supplies, independent subgroup firing for tonal control) and yielding only the expected result.

Explicit uncertainties I will not paper over:

1. Five of the fourteen of-record citations were not retrieved; claims 21–22 are truncated.
2. I have not obtained full texts of US 4,746,935, US 5,648,806, US 5,984,464, US 6,053,599, EP 0 897 804 A2, EP 0 440 490 B1, EP 0 913 257 A2, or JP H06-79873. Statements about their contents are based on titles, abstracts, and family/citation context and must be verified against the documents themselves.
3. My premise that the switched-return claims carry the 1999-04-27 effective date is an inference from the CIP structure and citation dates, not from inspection of the parent application files.
4. This is a technical analysis of invalidity risk; it is not a legal opinion, and it is offered against a patent that is already expired and has no located litigation history.
5. I did not locate any reexamination certificate, IPR, or PGR for '639; if one exists and narrowed or confirmed the claims, it would supersede this analysis.