US 5619236

I'll run targeted searches for this patent number, including litigation/docket checks.

Let me attempt to retrieve the abstract/claims text and check for any litigation records.

Let me try to retrieve the actual claim language from FreePatentsOnline / Justia.

I'll search for litigation involving US patent 5619236 across patent litigation databases.

Let me run more targeted searches on litigation databases.

Litigation Search Report — U.S. Patent No. 5,619,236

Patent identified literally: US 5,619,236 A — "Self-cooling printhead structure for inkjet printer with high density high frequency firing chambers"

• Application No. 08/648,471; filed 1996-05-15; granted 1997-04-08
• Priority date 1992-04-02 (claimed from US 07/862,086, which issued as US 5,278,584)
• Inventors: Keefe, Ho, Courian, Steinfield, Childers, Tappon, Trueba, Chapman, Knight, Moritz III
• Original assignee: Hewlett-Packard Company; current assignee: Hewlett-Packard Development Co. LP
• Legal status: Expired – Lifetime, anticipated expiration 2012-04-02

Bottom line

No litigation involving U.S. Patent No. 5,619,236 was located. I found no district court case, ITC Section 337 investigation, PTAB post-grant proceeding (IPR/PGR/CBM), or reported appellate decision in which the '236 patent is identified as an asserted patent.

What I searched

• General web searches for the patent number in litigation contexts ("5619236 litigation," "5619236 patent lawsuit," "5,619,236 infringement complaint")
• Targeted searches pairing the number with litigation aggregators (Docket Alarm, CourtListener/RECAP, PACER-style sources) and with the Unified Patents portal
• Assignee/technology searches (Hewlett-Packard inkjet printhead patent litigation)

Because my only available tool was general web search, I could not directly query the Unified Patents litigation caselist, PACER, or the CourtListener docket API. I therefore cannot positively exclude (a) sealed or otherwise unreported cases, (b) cases filed before roughly the mid-1990s that are poorly indexed online, or (c) cases in which the '236 patent was asserted but the number never appeared in an indexable public document. My conclusion is that no such litigation is publicly reported or indexed.

Critical caution: "similar numbers" results are NOT this patent

Per the strict-instruction rule, I am expressly flagging the many near-miss hits returned by search. None of the following involve US 5,619,236, and none should be reported as litigation of this patent:

Corroboration from the patent record

The Google Patents record for US 5,619,236 (https://patents.google.com/patent/US5619236/en) shows a prosecution/assignment history with no litigation or post-grant proceeding entries — only the standard assignment chain (Hewlett-Packard Company merger recorded 2001-01-16; assignment to Hewlett-Packard Development Company, L.P. recorded 2011-09-22) and the 2012-04-02 anticipated expiration. Google Patents generally surfaces U.S. district court litigation and PTAB proceedings in a case-event timeline when they exist; none appears here.

The patent also appears in the public record only as prior art cited against later filings — e.g., cited in the Russian patent RU 2 488 490 C2 (Canon) and as a family member of the HP printhead family listed against WO 1998/055316 A1 — not as a patent-in-suit.

Practical significance

The '236 patent expired on 2012-04-02 and its term predates the wave of post-2010 inkjet/printhead NPE assertion activity, which materially reduces the likelihood of any unfound assertion suit. If you need a negative-clearance–grade confirmation, the next step would be a direct PACER/RECAP docket-name search across all districts and a manual check of the Unified Patents litigation caselist and the USPTO PTAB API for the exact string "5,619,236" — none of which I was able to execute with the tools available here. I recommend that verification before relying on this as a definitive no-litigation opinion.

Verification trail (what I checked)

Number-substring trap — flag it explicitly: Japanese patent JP 5619236 B2 (Commissariat à l'Énergie Atomique, granted 2014-11-05, "connection component with hollow inserts") is a completely different patent. Any search hit or citation list conflating the two is a false positive. Likewise Brazilian/Lithuanian registry entries containing the string "5619236" are irrelevant.

Proceedings overview

Zero. US 5,619,236 has no AIA trial proceedings on file — no IPR, no PGR, no CBM, no derivation, at any point from the AIA's 2012-09-16 effective date through today (2026-10-01). Claim-level outcomes, panels, and FWDs therefore do not exist and I will not manufacture them.

The bottom-line defensive posture is unusual and, for a defendant, extremely favorable — but the reason is not that the patent survived challenges. It is that the patent is dead by operation of law: per the Google Patents record, US 5,619,236 was filed 1996-05-15 as a continuation of US 07/862,086 (priority 1992-04-02), issued 1997-04-08, and anticipated expiration 2012-04-02, with status "Expired – Lifetime." The absence of PTAB activity is a signature of expiration, not a signature of a hardened patent. The correct clinical description is "the patent has been in the public domain for ~14 years; the § 286 damages lookback closed on 2018-04-02" — not "the patent has survived two IPRs."

Proceedings

None to report. No petition has ever been filed against US 5,619,236 at the Board. Consequently there are no entries for type, filing date, panel, grounds, institution decisions, FWDs, settlements, or appeals. Any output that supplied an IPR number here would be fabricated.

For completeness on the related family (helpful context only, not a proceeding against this patent): the '236 patent claims priority to US 07/862,086, which issued as US 5,278,584, and it sits in HP's early-1990s edge-feed printhead family (siblings include US 5,563,642, 5,568,171, 5,594,481, 5,604,519, 5,635,966, 5,638,101, 5,648,804, 5,648,805, 5,648,806). Those family members were used by HP in worldwide cartridge enforcement programs in the 2010s (e.g., HP's 2013-01-23 Polish settlement press release; HP v. Benson Image B.V., ECLI:NL:RBDHA:2016:14326, Rechtbank Den Haag). I could not verify that 5,619,236 itself was among the asserted patents in any of those campaigns, and I will not state that it was.

Strategic summary

Claim status — CANCELED / SUSTAINED / UNTESTED. There is no PTAB record, so no claims are canceled or sustained by the Board. The operative claim status is statutory, not adjudicative: the entire patent — every claim, including the independent claims directed to the inkjet print cartridge with the edge-feed primary/secondary channel architecture and the 600 dpi first/second firing-chamber arrays recited in the summary — lapsed on 2012-04-02 for failure to pay maintenance / term expiry. As of 2026-10-01 every claim is expired and unenforceable. There is no "surviving claim" list to give you; the correct list is empty.

Estoppel landscape — § 315(e)(2). Not applicable in the direction you'd expect. Estoppel is petitioner-side, and there is no petitioner, so no IPR estoppel binds anyone. The flip side is that the patent owner's offensive options are foreclosed far more completely than any estoppel could achieve:

• Infringement damages: § 286 bars recovery for infringement more than six years before suit. The patent expired 2012-04-02, so the last day on which recoverable infringing activity could have occurred was 2012-04-02. Any § 286 recovery window closed 2018-04-02. A demand letter asserting this patent in 2026 cannot support a damages theory.
• Injunctive relief: impossible — there is no unexpired right to exclude. You cannot be enjoined from practicing subject matter in the public domain.
• IPR: technically the Board has instituted IPRs against expired patents in narrow circumstances (e.g., to clear a cloud while a parallel action or appeal is live). Here there is neither the practical benefit (you cannot be sued going forward) nor, in my view, a real prospect of institution where the patent has been expired for well over a decade. CBM is doubly unavailable — wrong subject matter, and the transitional program sunset 2020-09-16. PGR was never available (the patent issued 1997).
• Ex parte reexamination (§ 302): I flag this with only moderate confidence — the Office's practice on expired patents is that reexamination can be instituted but claim amendment is unavailable, which makes it a curiosity with no defensive value here. Do not budget for it.

Pattern signals. No repeat petitioner, no defensive aggregator (Unified Patents / RPX / Askeladden) in the chain, and no PTAB appeal history — because there is no chain. The only "pattern" is the ordinary one: a 1992-priority HP printhead patent that was cited as prior art by later artisans (e.g., the Canon/Kyocera-side search reports citing "US 5619236 A, 08.04.1997" in WO 2010/050959 A1 and RU 2 488 490 C2) and that quietly expired in 2012 without ever drawing a post-grant challenge.

Recommended next steps

1. If you received a demand letter or complaint asserting US 5,619,236, treat it as a red flag on the sender, not a threat. Confirm the expiration date on the face of the Google Patents record (Anticipated expiration 2012-04-02) and the damages cutoff (2018-04-02). A plaintiff asserting an expired patent faces Rule 11 exposure and, if suit was filed, potential § 285 fees. There is no FWD to link to here because no proceeding exists — the disposition you cite is the expiration, not a Board decision.
2. Do not accept a "hardened patent" narrative. If opposing counsel describes the patent as "never invalidated," that is technically true and legally meaningless: no one had a reason to challenge a patent that expired in 2012. Section 282's presumption of validity attaches to issued claims, but it confers no remedy on an expired patent.
3. If the assertion is instead against a family member (5,278,584, 5,563,642, 5,568,171, 5,594,481, 5,604,519, 5,635,966, 5,638,101, 5,648,804, 5,648,805 or 5,648,806 — all 1992-priority, all of which expired on or about 2012-2013), run the same expiration check first. These are contemporaneous, so they should all be expired; if a demand cites one, verify rather than assume, and ask me to run the same analysis on the specific number.
4. If a client asks whether to file an IPR on this number: no. No live exclusionary right, no damages window, statutory deadlines long past. Spend the budget on claim-charting the asserted family member or on the parallel district-court invalidity case.

Confidence statement. I have high confidence in the negative finding (no PTAB proceedings) — it is consistent across the structured ODP block and multiple search formulations. I have high confidence in the 2012-04-02 expiration date, sourced directly from the patent record. My statements about ex parte reexamination of expired patents and about the practical availability of IPRs against long-expired patents are moderate confidence and should be confirmed against current MPEP/Board practice if they become outcome-determinative.

I'll research the assignment record for US 5619236, starting with the patent's known legal events and then checking for any NPE-related activity.

Let me look for litigation history and any NPE activity involving this patent, plus the original priority application's assignment record.

I found a highly relevant precedent — the HP corporate merger reassignment pattern. Let me verify the specific reel/frame references and check for any NPE or litigation activity.

I have enough to reconstruct the chain. Note: the live Assignment Center UI could not be directly scraped through my search tool (it returned no direct record for 5619236), so I am grounding this on the Google Patents legal-events record (authoritative, reproduced in the patent text you supplied) plus sibling HP patents whose INPADOC assignment records expose the same corporate transfers. I flag inferred reel/frame numbers explicitly.

Inventors

All ten are consistent with a single captive corporate R&D team (HP's edge-feed thermal inkjet cartridge program, Corvallis OR / Palo Alto CA). The application was filed 1996-05-15 as US 08/648,471, a continuation of priority application US 07/862,086 (which issued as US 5,278,584), priority date 1992-04-02.

• Departure pattern: Not determinable from the assignment record. There is no evidence of a mass departure within 12 months of filing; the multi-inventor corporate-team signature is the opposite of the "inventors bail out, portfolio fire-sale follows" tell. No finding on this axis.

Original assignee

• Entity named on the issued patent: Hewlett-Packard Company (Palo Alto, CA; reincorporated in Delaware — later recorded as "Hewlett-Packard Company, Colorado").
• Product embodiment: Yes. The claims cover HP's edge-feed thermal inkjet printhead architecture (600 dpi, two offset resistor rows, self-cooling via ink flowing around the substrate edge). This family shipped in HP's DeskJet / OfficeJet cartridge line, so the original assignee practiced the invention.
• Primary line of business: Computing hardware, printers and imaging (HP was the dominant thermal-inkjet vendor in this era).
• Current status: Operating. HP remains a going concern (HP Inc. for the printing business). The patent is held within the HP corporate family via Hewlett-Packard Development Company, L.P. (Houston, TX), HP's IP-holding partnership. No bankruptcy, no dissolution. (The 2015 HP/HP Enterprise split did not move this patent out of the HP Development Company LP entity per the record below.)

Assignment timeline

Two recorded post-issuance assignments, both internal HP corporate reorganizations; no third-party or NPE transfer. (Google Patents legal events, reproduced in the authoritative patent text.)

• 1998-05-20 (executed/effective) / recorded 2001-01-16 — Reel 011523/0469 (reel/frame inferred from HP's batch-recorded sibling patents; see note below)

  • Conveyance: Merger
  • Assignor: Hewlett-Packard Company
  • Assignee: Hewlett-Packard Company (Colorado)
  • Correspondent: Hewlett-Packard Company, Intellectual Property Administration, Fort Collins, CO (HP's own in-house records department is the recurring correspondent on every HP assignment; see "sibling-record" note — this is an assignee-internal signature, not an outside NPE firm)
  • Context: Internal reorg — HP's 1998 Delaware reincorporation merger, batch-recorded across HP's portfolio in 2001. The patent never left HP ownership; the legal entity name changed.

• 2003-01-31 (executed/effective) / recorded 2011-09-22 — Reel 026945/0699 (reel/frame inferred from HP's batch-recorded sibling patents)

  • Conveyance: Assignment of Interest
  • Assignor: Hewlett-Packard Company
  • Assignee: Hewlett-Packard Development Company, L.P. (Houston, TX)
  • Correspondent: Hewlett-Packard Company, Intellectual Property Administration (same in-house correspondent)
  • Context: Internal reorg — migration of HP's patent portfolio into HP's dedicated IP-holding LP. Effective 2003 regardless of the 2011 recording date; the late recording is common for HP's bulk transfers.

• 2012-04-02 — Anticipated expiration (Google Patents legal event; not an assignment). Term ended; status "Expired – Lifetime."

Sibling-record note (why I flag the reel/frames as inferred): I could not pull the Assignment Center page for 5619236 directly. The reel/frame values above are taken from INPADOC records for other HP patents that went through the same two corporate events: US 5,185,860 shows the 2001-01-16 merger at REEL/FRAME 011523/0469 (effective 1998-05-20, "MERGER;ASSIGNOR:HEWLETT-PACKARD COMPANY"), and US 6,478,415 shows the HP→HP Development Company transfer at REEL/FRAME 026945/0699 (effective 2003-01-31). HP filed these as multi-thousand-patent batch assignments, so the same reel is very likely but not confirmed to be identical for 5619236. Verify at the Assignment Center link in the Verdict section.

If the premise of your question was "no records": there are records, but only the two HP corporate ones. This patent has no third-party assignment and no NPE transfer — it never left the HP family.

Timeline diagram

timeline
    title Ownership of US 5619236
    1992 : Priority application filed
    1996 : Application 08 648471 filed
    1997 : Patent issued to HP
    1998 : HP reincorporation merger effective
    2001 : Merger recorded reel 011523
    2003 : Transfer to HPDC effective
    2011 : HPDC transfer recorded reel 026945
    2012 : Patent expired

NPE / troll-pattern signals

Verdict

Operating-company assertion — with the explicit caveat that assertion itself is not evidenced. This is a 100%-HP-owned patent: the only recorded assignments are HP's 1998 reincorporation merger (reel 011523/0469, effective 1998-05-20) and the assignment of interest into HP's IP-holding LP (reel 026945/0699, effective 2003-01-31). No shell entity, no known asserter, no third-party or pre-litigation transfer, and the patent expired 2012-04-02 while still inside the HP family — so the NPE-pattern analysis is negative on every signal.

(Label caveat: because no suit enforcing this patent was found, the "assertion" prong of this verdict is unmet; the operative finding is simply non-NPE, original operating-company ownership throughout. If you require the assertion prong, the alternative closest label would be "insufficient data" for litigation — but the ownership record itself is complete and unambiguous.)

Verify:

• USPTO Assignment Center — search patent number 5619236: https://assignmentcenter.uspto.gov/
• Google Patents legal events: https://patents.google.com/patent/US5619236/en#legal-events

Do-not-fabricate disclosure: the exact reel/frame for 5619236 (011523/0469 and 026945/0699) is inferred from HP's batch-recorded sibling patents US 5,185,860 and US 6,478,415, not read off the 5619236 page directly. Everything else (inventor list, assignees, execution/recording dates, expiration) comes from the authoritative patent text and is not speculative.

I'll search for the specific patent's cited references and prior art.

Let me retrieve the specific "Patent Citations" (references cited) list from the patent's record.

Let me try to retrieve the front-page reference list directly from alternate patent databases.

Prior-Art / Citation Analysis — U.S. Patent No. 5,619,236

Note on method and a hard limitation, stated up front: My only available tool is general web search (no direct USPTO Patent Public Search / PatentCenter / PTAB API access). I searched repeatedly for the front‑page "References Cited" table of US 5,619,236 on Google Patents, FreePatentsOnline, and everypatent, plus EPO/EP and WO search‑report annexes that list the '236 family. I was not able to retrieve the actual examiner‑cited front‑page list (the U.S. Patent Documents / Foreign Patent Documents / Other Publications boxes) for US 5,619,236. Google Patents' crawler‑visible text for this record surfaced only the abstract, description, classification, and legal‑status data, and FPO/everypatent surfaced only description/claims text. Per your operating rules, I will not fabricate that citation list. Below I separate (A) what is verified, (B) what is verified but is not prior art, and (C) what still must be pulled from the USPTO record.

1. Document identity (literal, not auto‑corrected)

Distinct‑document warning (contradiction check vs. prior sections): "5619236" also resolves to JP 5619236 B2 (registered 2014‑09‑26, published 2014‑11‑05, Commissariat à l'énergie atomique, "接続構成部品" / hollow‑insert connection component). This is a different document in a separate register and is not this patent. This corroborates the earlier litigation section — no contradiction found.

2. (A) Verified citations appearing on the '236 record

The only references I could verify from the authoritative full text are those the specification itself cites:

A1. U.S. Patent No. 4,719,477

• Full citation: US 4,719,477 (Hewlett‑Packard)
• Status on '236: Cited in the specification and expressly incorporated by reference (the '236 text: "This development is described in U.S. Pat. Nos. 4,719,477 and 5,122,812 which are herein incorporated by reference.")
• Subject matter per the '236's own characterization: thermal inkjet printheads that incorporate pulse driver circuitry directly on the printhead substrate with the resistors, reducing interconnect count.
• Date: issued 1988 (pre‑dates the 1992‑04‑02 priority date) → available as §102(a)/(b) art and part of the '236 disclosure via incorporation.
• §102 relevance to '236 claims: Low / no anticipation of the independent claims. The '236 independent claims are directed to the edge‑feed architecture (firing chambers along a first outer edge, primary/secondary channel, separate inlet passage per chamber, 600 dpi chamber arrays, primitive grouping, >9 kHz firing). The '477 reference is relied on only for the on‑substrate driver circuitry concept, which appears in the '236 as background (the "first circuit on the substrate" / demultiplexer 78 element). At most it is relevant under §102 to any claim reciting only on‑substrate driver circuitry — not to the chamber/channel geometry claims.

A2. U.S. Patent No. 5,122,812

• Full citation: US 5,122,812 (Hewlett‑Packard)
• Status on '236: Cited in the specification and expressly incorporated by reference (same sentence as A1).
• Subject matter: same family of development — on‑substrate drive circuitry for thermal inkjet printheads.
• Date: issued 1992‑06‑16 — after the '236's 1992‑04‑02 priority date, so it is not available as §102(a) or §102(b) art; its only role is as incorporated subject matter. (Its own U.S. filing pre‑dates 1992, so a §102(e) theory would require the citation date and filing date to be checked against the record — see §5.)
• §102 relevance: Low, for the same reason as A1.

A3. Priority / parent document

• US 5,278,584 (issued 1994‑01‑11), from application 07/862,086 filed 1992‑04‑02 — the '236's claimed priority parent.
• Not prior art. Common ownership + same inventors + same priority chain. It is, however, the single most likely entry to appear on the '236 front page as a related document, and it is frequently mis‑reported as prior art against the '236.

3. (B) Verified family / reverse‑citation documents — NOT prior art against '236

These are well documented in the EPO and PCT search‑report annexes I retrieved. They share the 1992‑04‑02 / EP 0 564 069 / CA 2 083 341 / JP 06‑008434 priority and are family members, not references cited against '236:

US 5,278,584 · US 5,434,607 · US 5,563,642 · US 5,568,171 · US 5,594,481 · US 5,604,519 · US 5,625,386 · US 5,635,966 · US 5,638,101 · US 5,648,804 · US 5,648,805 · US 5,648,806 · US 5,873,974 · US 5,946,012 · US 5,953,029 · US 5,984,464 · US 6,183,076 · US 6,332,677 — with EP 0 564 069 A2/B1, EP 0 705 693 A2/B1, CA 2 083 341 A1/C, DE 69305401 D1/T2, ES 2093359 T3, HK 92997 A, JP 3386177 B2, JP 3699175 B2, KR 224952 B1.

Documents that cite the '236' (forward citations) — the inverse direction, explicitly not prior art:

• WO 2010/050959 A1 — cites "US 5619236 A (KEEFE et al.) 8 April 1997 … See abstract, figure 13 and column 10, line 56 – column 11, line 65," Category A (technological background only).
• RU 2 488 490 C2 (Canon) — lists US 5619236 A (08.04.1997) in its search‑report citation list.
• US 5,568,171 (Keefe et al.), cited in WO 1998/055316 A1 and WO 2009/018316 A1, is a sibling, not a reference against '236.
• US 7,407,271 (Silverbrook) cites US 5,619,236 among its "Referenced Cited" U.S. patent documents — again a forward citation.

Anyone who treats any of the above as "prior art for 5619236" has reversed the arrow.

4. (C) What I could NOT verify — the actual examiner‑cited list

I could not retrieve the front‑page boxes for US 5,619,236. Consequently:

• I cannot give you a per‑reference §102 mapping against an examiner‑cited list, because I do not have that list.
• I will not reproduce a remembered list of 1980s/90s HP/Kodak/Canon inkjet patents (e.g., early thermal‑inkjet "edge‑shooter" or "hydraulically tuned channel" references, US 4,490,728, US 4,882,595, US 4,417,251, US 4,532,530, etc.) and present them as citations of '236. Several of those are plausible candidates by subject matter, but I have no evidence they appear on the '236's face, and the only search hit relating to US 4,882,595 ("Hydraulically tuned channel architecture," cross‑talk/meniscus control) came from my own query string, not from the '236 record — so it is not evidentiary.

To close this gap, these four pulls are required:

1. USPTO Patent Public Search → US5619236 → "Referenced By / References Cited" (the 1449‑sourced boxes).
2. PatentCenter → Application 08/648,471 → the PTO‑1449 / IDS and the file wrapper (any examiner‑cited art from the parent 07/862,086).
3. Google Patents → US5619236 → "Patent Citations" and "Cited By" tables (render via the direct anchor #patentCitations, which my crawl did not expose).
4. EPO Global Dossier / Espacenet → family entry for EP 0 564 069 → the EPO search report (the European counterpart's cited art frequently overlaps the U.S. front page).

5. §102 framework you should apply once the list is in hand

Critical dates for the '236 (pre‑AIA statute):

Anticipation analysis of the verified references:

• US 4,719,477 — disposes of, at most, an on‑substrate driver‑circuitry limitation. It does not disclose the '236 independent‑claim combination: (i) firing chambers along a first outer edge of the substrate with the edge in close proximity to the chambers; (ii) a primary channel conducting ink from the reservoir around the substrate edge to a secondary channel on the top surface; (iii) a separate inlet passage per chamber for high‑frequency refill; (iv) first and second chamber arrays spaced for 600 dpi; (v) a primitive group in which only one chamber fires at a time; (vi) firing signals at >9 kHz. No §102 anticipation of claims 1/independent claims.
• US 5,122,812 — issued 1992‑06‑16, i.e. post‑priority; unavailable under §102(a)/(b). §102(e) only if its application was filed before 1992‑04‑02; verify on the record before asserting. Same disclosure‑scope limitation as above.
• US 5,278,584 (parent) — excluded as prior art by common priority; usable only as an obviousness/claim‑construction cross‑reference.

Bottom line on anticipation: On the material I could verify, no reference anticipates the '236's independent claims. The novelty of the '236 sits in the edge‑feed barrier architecture (specifically the convergence of: short "shelf length" U, pinch‑point gaps 145/146 for viscous damping, peninsulas 149 and reefs 148, 600 dpi two‑row chamber pitch with offset E, and the primitive/>9 kHz drive scheme) — see the '236 description at the FIG. 15–18 and FIG. 20–23 passages, and the abstract. The references I can verify address only the on‑substrate drive‑circuitry aspect.

6. Consistency check against the previously generated sections

• The earlier Litigation summary concluded no litigation on US 5,619,236 and flagged the JP 5619236 / US 7,669,236 / US 6,516,236 / US 7,470,236 near‑misses. No contradiction with anything I found; my searches independently reproduce the JP 5619236 B2 document and the reverse‑citation (Category "A") treatment of US 5619236 A in WO 2010/050959 A1 and RU 2 488 490 C2.
• The earlier Patent summary did not enumerate the front‑page citations either — consistent with the retrieval limitation I am documenting here.
• One clarification to add rather than contradict: the '236 is a continuation of 07/862,086 (US 5,278,584), so US 5,278,584 and the eleven sibling U.S. patents sharing EP 0 564 069 are family members and forward/related documents, never §102 art. The earlier sections did not distinguish these from cited art; doing so here.

Honest bottom line

I verified only two citations on the '236 record — US 4,719,477 and US 5,122,812, both incorporated by reference for on‑substrate driver circuitry — plus the parent US 5,278,584 and the extensive EP 0 564 069 family. Neither verified reference anticipates the '236's independent claims. I could not retrieve the examiner‑cited front‑page reference list with the tools available, and I have deliberately declined to reconstruct it from memory, because presenting unverified numbers as citations of 5619236 would violate the strict‑literal‑identifier rule. Supply the 1449/file‑wrapper list (or the Google Patents #patentCitations table) and I will produce the per‑reference §102 mapping you asked for.

I'll pull the prior-art references actually cited in US 5,619,236 and its family, then build the § 103 combinations.

The full text I was given is truncated before the claims. Let me pull the actual claim set and the front-page (56) references for the '236 patent.

Obviousness Analysis Under 35 U.S.C. § 103 — U.S. Patent No. 5,619,236

Builds on the Patent Summary and Litigation Summary already generated. This section does not repeat the bibliographic record (app. 08/648,471; filed 1996-05-15; granted 1997-04-08; priority 1992-04-02 from 07/862,086 → US 5,278,584) except where the § 103 analysis depends on it.

1. Threshold caveats — read before relying on any ground below

The source text I was given is truncated mid-specification at FIG. 20 / "Preferred dimensions A, B, and C … Table IV." It does not contain the claims. I also could not retrieve the '236 front page's (56) "References Cited" list with the tools available; the three targeted fetches for the claim set and the (56) list were cut off by the tool-step limit.

Therefore:

• This analysis is keyed to (a) the abstract, (b) the "Definitions"/summary passages in the specification I was given, and (c) the disclosure of the sibling family members I could retrieve. It is not a claim-by-claim § 103 opinion. Treat every "claim element" below as a probable element pending retrieval of the actual claim text.
• Every reference I identify is labeled with how I verified it (cited inside the '236/family specification; retrieved as a sibling; found in a search report). I have flagged as unverified any reference I could not confirm appears on the '236 front page.
• Patent numbers are taken literally. In particular, JP 5619236 B2 (Commissariat à l'énergie atomique, hollow insert connection component) is a different register and different invention; it is not the US patent analyzed here. This was already flagged in the Litigation Summary.

2. The pivotal issue: effective filing date, and which art is actually available

This is where the § 103 analysis turns, and it is worth being blunt about it.

The '236 is a member of a large HP continuation family radiating from application 07/862,086 (1992-04-02, issued as US 5,278,584). My searches retrieved the front matter of two siblings:

• US 5,648,805 (Keefe et al.), "Inkjet printhead architecture for high speed and high resolution printing," Appl. No. 319,896, filed Oct. 6, 1994, a continuation-in-part of Ser. No. 179,866 (Jan. 11, 1994 — itself a continuation of 862,086), with the notice: "The term of this patent shall not extend beyond the expiration date of Pat. No. 5,278,584."
• US 5,625,396, expressly "a continuation of U.S. application Ser. No. 07/862,086 filed Apr. 2, 1992."

(Mirror: https://patentimages.storage.googleapis.com/dd/8b/df/56b4e40d1005ab/US5625396.pdf and …/7e/65/8d/7b9d939f73cd57/US5635966.pdf)

Why this matters. The features that make the '236 commercially interesting — 600 dpi, a primitive in which only one firing chamber is activated at a time, and firing signals at a frequency greater than 9 kHz (indeed the '805 abstract recites up to 12 KHz) — map onto the 1994 CIP disclosure (US 5,648,805), not obviously onto the 1992 parent (the '584 material I retrieved describes 300 nozzles / 300 dpi). Consequences:

Critically, the siblings are not § 102(e)/(a) prior art against the '236. US 5,648,805, US 5,625,396, US 5,635,966 are by the same inventive entity (Keefe, Ho, Courian, Steinfield, Childers, Tappon, Trueba, Chapman, Knight, Moritz) — 102(e) requires a reference "by another." They are also commonly owned by HP, so even if they qualified as 102(e) art, pre-AIA § 103(c) would disqualify them for obviousness. They are therefore useless as § 103 references for affirmatively invalidating the '236, even though they disclose every disputed feature. That is a real-world limitation on any invalidity theory, and it should be stated plainly: the most damaging-looking documents in this family are unavailable.

US 5,278,584 itself is different: it issued 1994-01-11, more than one year before the 1996-05-15 filing, so as a patent it is 102(b) art on its face. But it is the parent disclosure; claims fully supported by 1992 matter are then a § 102(b) / § 120–121 issue, not a § 103 combination. That distinction should be preserved rather than blurred.

3. Prior-art inventory (verified vs. unverified)

Post-dated material I deliberately exclude as non-prior-art: US 5,850,234 (ink-inlet design for increased ink cooling of the printhead) and JP H10-157155 are after the '236's filing; they are useful only to show that ink-as-coolant was routine, and I flag them as such.

4. Grounds of rejection

Ground 1 — P2 (Johnson) + P3 (Buck) + P6 (Hess) + P11 (Kodak-type cooling)

Johnson supplies the substrate/barrier-layer/chamber/resistor/nozzle stack; Buck supplies the disposable thin-film cartridge environment; Hess supplies drive circuitry fabricated on the printhead substrate with a demultiplexer so that many resistors can be addressed with few interconnects; Kodak-type art supplies fluid-based heat removal from the printhead.

Motivation: the '236's own background frames the problem identically — "prior printhead design limited the ability of printheads to have the high nozzle densities and the high operating frequencies and firing rates required for increased resolution and throughput," and "adding them … increases the cost and reduces the reliability of the printer" for interconnections. Hess's on-substrate addressing is the recognized answer to that exact interconnection problem, and is expressly incorporated by reference into the '236.

Failure of this ground: it does not reach (i) ink flowing around the outer edge of the substrate to a secondary channel on the top surface, (ii) the separate per-chamber inlet passage sized for high-frequency refill, (iii) 600 dpi in first/second arrays, (iv) the one-at-a-time primitive, or (v) >9 kHz. Johnson teaches away from the claimed geometry — it is a center-feed manifold, which the '236 criticizes for flow restriction and crosstalk. Ground 1 alone is insufficient. It is included because it defines the baseline the edge-feed geometry departed from.

Ground 2 — P1 (US 5,278,584) + P10 (Bhaskar offset heater resistors) + P6 (Hess)

This is the strongest combination, and it is the one an examiner would most plausibly have articulated:

Motivation to combine, stated in KSR terms: (1) the field is the same — thermal inkjet print cartridges; (2) the '236 advertises the known problems — "the closer spacing of the resistors created space problems," "cross-talk between neighboring vaporization chambers becomes a serious problem," and "an increased number of heater resistors spaced closer together and higher resistor firing frequencies increases the heat which must be dissipated"; (3) the asserted solutions are recognized, incremental countermeasures to those known problems — stagger the emitters, fire only one per primitive, shorten the shelf, pinch the channel to damp the meniscus; (4) the results (denser dots, faster refill, less crosstalk) are predictable in direction and magnitude; (5) there is a clear market pull to 600 dpi, recited as a numerical target rather than a discovered principle. Where a claim recites a result-effective parameter such as "greater than 9 kHz" or "600 dpi," the recitation is at most a design target, not a technical contribution, absent a showing that the specific value produced an unexpected effect.

Ground 3 — The "self-cooling" limitation

If any claim recites heat dissipation by the ink, that is an inherent property of the P1 edge-feed geometry: ink traverses the back surface and the peripheral edges on its way to the chambers. The '236 specification says so itself — "the substrate is able to dissipate more heat, since the ink flowing across the back of the substrate and around the edges of the substrate acts to draw heat away." Discovering an advantageous, inherent behavior of an otherwise-known structure, without more, is not a patentable distinction; and fluid cooling of a printhead was itself known (P11 class). This feature is weak as a saving limitation unless the claim couples it to a structural change (e.g., a specific shelf length U or pinch-point gap) that produced a genuinely unexpected result.

Ground 4 — The pinch-point / shelf-length damping limitations (FIGS. 16–17)

The specification concedes these are tuning choices: "as the shelf length U decreases in length, the nozzle frequency increases," and the pinch point is "intentionally controlled." A recitation of "a well-defined length and width" constriction to damp meniscus oscillation, with the trade-off between impedance and damping expressly acknowledged, is the classic case of an optimization of a known parameter with a predictable result — obviousness territory. But note the counter-argument: the combination of a shortened shelf (reduced moving-ink mass), a pinched channel, and offset emitters, applied to sub-2-mil barrier layers at 600 dpi, could be argued to be a non-obvious interaction rather than an aggregation. I would expect that argument to be the patentee's best § 103 rebuttal, and it is at least colorable.

5. Secondary considerations (Graham factors)

• Unexpected results: the only candidate is the self-cooling/heat-dissipation benefit. Weak nexus unless claimed, and the specification frames it as expected. The '236's own background presents edge feed's advantages as improvements over the prior center-feed design — i.e., as foreseen advantages of a design choice, not surprises.
• Teaching away: the strongest available argument. Johnson-type center-feed art taught routing ink through a central slot; the '236's criticism that center feed restricts flow, causes crosstalk, enlarges the die and makes it fragile ("once the hole is formed, the substrate is relatively fragile, making handling more difficult") is a reason to leave center feed, and one could frame the prior art as pointing to center feed as the standard. However, P1 already taught the departure; teaching away from the reference is not teaching away from the claimed combination.
• Long-felt need / commercial success: plausibly present (HP's high-resolution pens), but I have no evidence of a nexus between any specific claim and the commercial result, and I will not assert one.

6. Honest assessment

• Most likely outcome: the broad, geometry-only claims (reservoir + edge feed + primary/secondary channel + per-chamber inlet + two arrays) are vulnerable under § 103 over P1 in view of P10 and P6, with the density and frequency numbers treated as results. The claims most likely to survive are those that tie the high density to a specific structural interaction — the shortened shelf U combined with the pinch-point gap 145 combined with offset emitters — where a genuine, documented trade-off is resolved.
• What I could not verify and would need before signing this: (1) the actual claim text; (2) the '236 front page (56) list; (3) the priority determination as between the 1992, 1994 and 1996 disclosures; (4) the content of P3 (Buck), P4, P5, P8, P9 and P10, none of which I read directly; (5) whether any of the search hits at P11–P14 are in fact front-page citations of the '236 rather than mere co-occurrence in search results.
• Note a possible contradiction with the prior sections: the Patent Summary/Litigation Summary treated the '236 as a straightforward continuation of 07/862,086. The sibling record I retrieved (US 5,648,805, a CIP filed 1994-10-06, and US 5,625,396, a continuation of 07/862,086) shows the family contains both continuation and CIP branches, and the '236's own abstract recites 600 dpi / >9 kHz subject matter that the 1992 disclosure does not obviously support. That is a genuine tension with the "priority 1992-04-02" framing, and it materially affects which references qualify as prior art. It should be resolved by pulling the '236's [63] "Related U.S. Application Data" and the '584 patent's full specification.

Suggested next steps: retrieve the '236 claims and front page via USPTO PatentCenter / Patent Public Search for application 08/648,471; pull the [63] continuity data for the '236 and for US 5,648,805 to fix the effective date; then re-run the element map in § 4 against the literal claim language.