US 3996429

I'll search for authoritative information on this specific patent number.

Let me run additional targeted searches for litigation and docket records.

US Patent 3,996,429 — Analyst Summary

Search coverage / confidence note: I searched for the identifier as written (3996429 / 3,996,429). I found no CAFC 2026 docket entry and no district-court or PTAB litigation record naming this patent. Note also that the numeral string "3996429" appears in unrelated registries (a Belgian/EU patent-validation publication and a Lithuanian legal-notice serial), which are not this US patent. The patent is long expired, so absence of 2026 litigation activity is expected rather than surprising.

Bibliographic data

Source: https://patents.google.com/patent/US3996429/en and https://uspto.report/patent/grant/3996429

Abstract (as published)

"A push-button switch has a spring contact plate and a circuit board in superposed position. The spring contact plate has one or more switch positions, each switch position comprising a prestressed domed portion surrounded by a common flat sheet portion. The spring contacts of each prestressed dome portion selectively engage peripheral contact areas on the circuit board followed by the central contact portion engaging subsequently with a further fixed central contact area on the circuit board. Accordingly, plural inputs of electronic logic circuits may be preselected prior to simultaneous activation upon subsequent closure of the central contacts at a given switch position. Release of the push button enables the domed portion to snap back to the stable position."

Plain-language overview of the claims

Independent claim 1 — the multi-contact snap-action pushbutton switch.
A single flat spring plate (e.g., a stainless-steel stamping) carries several integral switch positions. At each position:

• A prestressed domed portion bulges upward out of the flat plate and is completely surrounded by the flat plate, so the surrounding flat material circumferentially constrains the dome.
• The dome has a center portion joined to the surrounding flat plate by several radial webs. Each web is upwardly bent where it meets the flat plate, and the webs are radially constrained by the flat portion.
• A central contact sits on the center portion.
• Cantilevered spring contact members project from the center portion, one between each pair of radial webs, reaching a shorter radial distance than the webs, angling outward and downward, each ending in a contact area. That contact area is positioned below the central contact.
• Below and spaced from the plate is a circuit member (the printed circuit board) with contact areas under the spring contacts, plus a further contact area under each central contact.
• A pushbutton sits above each dome. Pressing down makes the dome snap toward the circuit board and makes the webs bend just inside the upward bend, which generates a return force that snaps the dome back on release.
• Key functional limitation: because the spring-contact contact areas extend below the central contact, they touch the board's peripheral contact areas first, and only afterwards does the central contact close against its own contact area. So the code-select contacts are all pre-closed (pre-selected) and then simultaneously "gated" by the delayed common/central contact.

Dependent claim 2 — packaging detail. Adds a cover member over the spring contact plate with one aperture per pushbutton, plus interengaging formations on the cover and buttons that retain the buttons in place (the flange-in-recess arrangement described at button 35/flange 36/recess 33).

Dependent claim 3 — binary-code application. Narrows the switch of claim 2 to a binary coded signal system, in which the contact areas are arranged in a predetermined contact configuration so that depressing a given button produces a predetermined binary code (the specification's Table I gives codes for a 2 + 5 ten-button array, and FIG. 11 shows four 16-button modules multiplexed to 64 keys).

Technical significance noted in the specification

The delayed central contact is a common contact in series with the encoding contacts. This eliminates the need for a time-delayed strobe/clock signal to read the coded data and eliminates encoding interface logic, giving (per the patent) guaranteed simultaneous appearance of encoding bits, lower cost, and simpler mechanical construction. The patents also provide a strobe output for the "all-zero" (button 0) case and a grounded/resistor path to the spring plate to drain static charge and suppress arcing on contact break.

Additions to the prior sections

• Closest same-family/same-priority relative: US 4,029,916 (Northern Electric Company Limited, priority 1975-04-18, issued 1977-06-14), Multi-contact push-button switch and plural embodiment for keyboard switch assembly — same inventors' subject matter and same priority date, and the most relevant citation-linked document. Note its description differs in detail (it describes an elastic diaphragm over the contact plate and a contact board rather than the printed circuit board/cover arrangement of 3,996,429), which I flag because the two documents are easy to conflate.
• Cited prior art: 12 US references, including US 3,133,170 (Nanninga), US 3,600,528 (Lematex), US 3,741,957 (Lematex), US 3,699,294 (Flex Key), US 3,761,944 (Alps), US 3,941,953 (Hewlett-Packard), US 3,826,882 (Olivetti), US 3,898,421 (Matsushita), US 3,870,840, US 3,800,104, US 3,796,843 (Bowmar), US 3,941,964 (Bowmar).
• Forward citations: roughly 66–70 documents, including EP 0 084 734 A1 and EP 1 087 416 A1 (Molex) which cite it as an [A]/[XA] reference, and US 4,486,637 (Northern Telecom), indicating the art family remained active commercially into the 1980s.

Statement of uncertainty

• I have no authoritative information on any 2026 CAFC appeal, IPR/PGR, or district-court case involving US 3,996,429. My searches returned none, which is consistent with the patent having expired in 1993 but does not by itself prove no record exists.
• The current-assignee entry ("Nortel Networks Ltd") is a Google Patents assumption, expressly disclaimed by the source; the original assignee (Northern Electric Company Limited) is the reliable figure from the front page.

I'll run targeted litigation searches for this specific patent number.

Let me search litigation-specific databases and related terms.

Litigation search result: US 3,996,429

Finding: No litigation involving US patent 3,996,429 was found. I could locate no district-court action, no PTAB proceeding (IPR/PGR/CBM), no ITC Section 337 investigation, and no CAFC appeal naming this patent.

I searched for the identifier as written — 3996429 and 3,996,429 — including patent-litigation-oriented sources (Unified Patents litigation/PTAB case material, Docket Alarm PTAB dockets, USPTO PTACTS/PTAB petition records) and general web indexes. I found no case record.

Why the null result is expected, not merely a search gap

The "do not auto-correct identifiers" caution is directly relevant here. The numeral string 3996429 is not unique to this US patent, and my searches surfaced several false positives that must not be attributed to US 3,996,429:

• A Belgian/EU patent-validation publication (economie.fgov.be, Recueil brevets) listing (11) 3996429 against an LG Electronics wireless-communication application — a European validation number, unrelated to this US patent.
• A numerical coincidental match in an academic mixed-integer-programming paper (a solution value "0.3996429" appearing in the text).
• Various 3996429 hits in unrelated legal-notice and mathematical contexts.

None of these is a US patent case. Conversely, the genuine patent does appear in the literature as cited prior art — e.g., EP 1 087 416 A1 (Molex) cites "US 3 996 429 A (CHU PAK-JONG ET AL)" as an X-category reference against its claims 1 and 12, and TW I666669 B (Darfon Electronics, 2019) lists US 3996429 as a reference. Citation in an office action is not litigation, and I flag it here only so it is not mistaken for an enforcement event.

A bounded caveat on completeness

Patent-litigation records for the 1975–1993 window are not uniformly digitized. Older docket sheets and regional-district filings from the pre-CM/ECF era are frequently absent from the free databases I can reach, and a suit filed by Northern Electric Company Limited / Bell-Northern Research against a keyboard or telephone-set maker in that period would not necessarily be indexed by patent number in modern portals. So the correct statement is:

• As of 2026-04-26, no litigation involving US 3,996,429 is known or retrievable.
• I have no authoritative confirmation that no such suit ever existed; I can only report that none appears in the sources searched.

Given the patent has been expired since 1993-12-07, an IPR/PGR, a 2026 infringement action, or a 2026 CAFC appeal naming it are all legally impossible — the PGR window closed long ago, IPR is available only for unexpired patents (though the Board retains jurisdiction over already-instituted proceedings after expiration), and a CAFC appeal requires an underlying live case. Any 2026-date hit on the bare numeral 3996429 should therefore be presumed to be one of the identifier collisions above until verified on the face of the document.

Reconciliation with the previously generated summary

This task's result is consistent with the prior section, which likewise reported no district-court or PTAB record. No contradiction to flag. One date note for the record: the prior section was generated under a reported date of 2026-09-30, whereas this task states the current date as April 26, 2026; the system-supplied "patent full text" fetch timestamp also reads 2026-09-30. I have used April 26, 2026 as directed for this task. The ~5-month discrepancy does not change the analysis, since the patent's 1993 expiration is dispositive either way.

Recommended next steps if a definitive negative is required

1. USPTO Patent Assignment Search / Patent Center — check for any recorded nunc pro tunc assignment or license that would imply enforcement.
2. Docket Alarm, Lex Machina, or Westlaw Litigation Analytics — these have deeper historical coverage than the free portals I searched; query on patent number 3996429 and on assignee strings "Northern Electric Company Limited," "Northern Telecom," and "Nortel" separately.
3. Federal Records Centers / district-court archives for the pre-ECF era, if a specific suspected defendant period is of interest.

I do not have high confidence that any further free-source search will produce a hit, and I will not speculate about case names, case numbers, or filing dates that I could not verify.

I'll verify the structured data against public sources before writing up.

Let me run a couple more targeted checks for any record of an AIA trial or appeal tied to this patent.

One more check for any ex parte/inter partes reexamination or Federal Circuit record.

Proceedings overview

Total AIA trial proceedings on file for US 3,996,429: zero. The structured "PTAB proceedings on file" block (USPTO Open Data Portal, canonical source) returns no IPR, PGR, or CBM involving this patent, and my web searches returned no proceeding numbers, no institution decisions, no Final Written Decisions, and no Federal Circuit appeal — so there is nothing in the breakdown categories (active / invalidated / sustained / settled / institution denied) to report: the counts are 0 / 0 / 0 / 0 / 0. Bottom line for a defendant: there is no PTAB record to lean on and none to fear — but more importantly, none is needed. This patent expired 1993-12-07, roughly nineteen years before the first IPR was ever filed, so the absence of PTAB activity is structural, not a sign that the patent has "survived" anything.

One housekeeping flag so the record is not misread: the Google Patents page for US 3,996,429 lists 12 "Citations" (prosecution prior art) and roughly 66–70 "Cited By" entries. Those are prior-art citations made by later patents and examiners, not contested proceedings. Two of them are directly relevant as art, not as trials: EP 1 087 416 A1 (Molex) cites US 3,996,429 A as an X-category reference against its claims 1 and 12, and TWI666669B (Darfon Electronics) lists it in its search report. Similarly, the numeral string "3996429" turns up in unrelated registries (a Belgian FPS Economy / European Patent Validation publication number and a numeric result in a mixed-integer-programming paper) — neither is this US patent. See https://patents.google.com/patent/US3996429/en and https://uspto.report/patent/grant/3996429.

No proceedings to itemize

Because there are zero proceedings, I am not going to manufacture placeholder entries with invented proceeding numbers — that would be exactly the fabrication the constraints prohibit. For completeness, here is what I affirmatively checked and did not find:

Why the door is effectively closed, even though it is not literally bolted:

• PGR is unavailable. Post-grant review applies only to patents subject to the first-inventor-to-file provisions; US 3,996,429 has a 1975-04-18 priority date and is a pre-AIA patent.
• CBM is unavailable. The transitional covered-business-method program ran only from 2012-09-16 and sunset on 2020-09-16; the patent is also not a "covered business method" patent (it claims a pushbutton switch, a technological invention).
• IPR is technically still available. Under § 311(c) and the AIA's applicability provision, IPR reaches "any patent issued before, on, or after" 2012-09-16, and the only petition-filing gate (9 months after grant) lapsed in 1977. So a petition is not statutorily time-barred. But there is no economically rational reason to file one, for the reasons in the strategic summary below.

Defensive value of the empty record

If someone asserts US 3,996,429 against you today, the operative defense is not an IPR — it is the calendar. The patent's 17-year term from its 1976-12-07 issue date ran to 1993-12-07 (the Google Patents "anticipated expiration" field), and no later maintenance-fee-driven extension changes that. Infringement liability for a product shipped after 1993-12-07 simply does not exist, and the six-year damages lookback under 35 U.S.C. § 286 closed around 1999. A demand letter citing this patent is not a validity contest; it is a candidate for a fee-shifting or § 285 "exceptional case" response.

Strategic summary

Claim status: untested, not canceled, not sustained. Claims 1, 2, and 3 (claim 1 independent; claims 2 and 3 dependent) have never been challenged in any AIA trial, and I found no reexamination certificate altering their text. So "CANCELED vs. SUSTAINED vs. UNTESTED" resolves entirely to UNTESTED — 3 of 3 claims. This is the opposite of the usual defensive narrative: there is no FWD to point at, no claim-level disposition to quote, and no Board reasoning to borrow. Do not let anyone characterize the empty PTAB record as "the patent has been validated" — an unchallenged expired patent has been validated by no one.

Estoppel landscape: essentially vacuous, and that is the good news. Section 315(e)(2) estoppel is petitioner-specific — it attaches only after a petitioner obtains an FWD. With no petitioners, no FWDs, and no institution decisions, no estoppel has attached to any party, and no printed-publication or patent-art ground is foreclosed to a defendant by reason of § 315(e)(2). (Section 325(e)(1)/(2) estoppel is likewise inapplicable — no PGR ever existed.) Practically, though, estoppel is the wrong frame here: the prior art cited during prosecution (US 3,133,170 Nanninga; US 3,600,528 and US 3,741,957 Lematex; US 3,699,294 Flex Key; US 3,761,944 Alps; US 3,941,953 Hewlett-Packard; US 3,826,882 Olivetti; US 3,898,421 Matsushita; US 3,870,840; US 3,800,104; US 3,796,843 Bowmar; US 3,941,964 Bowmar) remains fully available for an invalidity defense in district court under §§ 102/103, unconstrained by IPR claim-scope limits or the Board's institution standards.

Pattern signals: none of the usual ones are present. There is no serial petitioner (no petitioner at all), no patent-owner PTAB appeal practice to reverse-engineer (no FWD ever issued), and no defensive aggregator in the chain — Unified Patents and similar entities do not appear anywhere in this patent's history, which is unsurprising given that it expired in 1993, about twenty-one years before Unified's formation. The only entity-interest fact worth noting is the current-assignee field ("Nortel Networks Ltd," successor to original assignee Northern Electric Company Limited), which Google Patents expressly disclaims as an assumption and which I could not verify from a primary USPTO assignment record. Nortel's bankruptcy and the 2011 Rockstar/Nortel patent-sale process did not, on the record I can find, put this patent into any assertion campaign — consistent with its expired status. The nearest family member, US 4,029,916 (Northern Electric, priority 1975-04-18, issued 1977-06-14), is likewise a citation relationship, not a proceeding.

Recommended next steps

• No PTAB activity exists on this patent — say so plainly and stop looking for an IPR-based defense. There is no FWD to link to, no institution decision to distinguish, and no proceeding number to cite. I searched the structured ODP block and public sources for IPR/PGR/CBM numbers, ex parte and inter partes reexamination certificates, and Federal Circuit docket entries; all came back empty. The absence is not the usual "well-asserted patents eventually attract IPRs" signal, because the patent expired 1993-12-07 — nearly two decades before the AIA created these forums. Any IPR-like filing today would be a solution in search of a problem.

• Treat the expiration date as the dispositive defense, and frame it that way in writing. Confirm the expiration and the absence of any post-expiration damages window against the primary records: the Google Patents front page ("Anticipated expiration 1993-12-07"; legal status "Expired – Lifetime") at https://patents.google.com/patent/US3996429/en, and the PTAB's own public docket tools at https://ptacts.uspto.gov (PTAB E2E / PTAB Center) if you want a fresh, dated screenshot for a file. A § 285 motion, a Rule 11 safe-harbor letter, or a state-law/declaratory-judgment posture are the realistic tools here — not an IPR.

• If a demand letter arrives citing this patent number, verify the identifier before responding on the merits. The numeral string circulates in unrelated registries (a Belgian/EU patent-validation publication number and a mathematical-paper value), and the patent's own Google Patents "Cited By" list is easy to misread as a litigation history. Confirm the letter is actually invoking US 3,996,429 and not a similarly numbered or family-adjacent patent before you build a response.

• Preserve the invalidity record anyway, cheaply. Should any party later attempt to assert this patent (or a family member such as US 4,029,916) in a non-US forum where a counterpart exists, the twelve prosecution references listed above and the Google Patents citation record are a ready-made § 102/103 package requiring no PTAB estoppel analysis at all.

Confidence and uncertainty statement: I have high confidence that no AIA trial proceeding exists for US 3,996,429, based on the canonical ODP-derived block plus corroborating searches. I have high confidence the patent expired 1993-12-07, as stated on the Google Patents front page. I have lower confidence on the current-assignee chain, which rests on a source that disclaims its own accuracy. I could not exhaustively query CourtListener or PACER dockets directly, so I cannot prove the negative on every conceivable peripheral filing — but the absence of any PTAB activity, any reexamination certificate, and any appellate opinion is consistent across every source I checked.

I'll research the assignment record for this patent.

Let me dig deeper into the specific assignment records and corporate history.

Let me check for any recorded assignments and the Nortel corporate/bankruptcy history.

Ownership Analysis — US Patent 3,996,429

Data-access caveat up front (read before the sections below). I was unable to retrieve any recorded assignment from the USPTO Assignment Center for this patent, and I could not execute a direct query against assignmentcenter.uspto.gov. I therefore report a retrieval failure / no-records-found, not a verified affirmative "zero records." Two structural facts frame everything below and are the most important findings in this report:

1. The USPTO's online Patent Assignment Search only covers records from August 1980 forward. MPEP 301.01 states the historical assignment database is "maintained separate and apart from the database of assignments affecting title recorded since 1980," and the USPTO's own Assignment Search documentation confirms coverage "back to August 1980." Source: https://www.uspto.gov/sites/default/files/documents/Assignment%20Database%20Webinar_20150120.pdf and https://blueironip.com/ufaqs/how-are-historical-patent-assignment-records-maintained/. This patent was filed 1975-04-18 and issued 1976-12-07, so its only probable assignment — inventors → Northern Electric Company Limited — would have been recorded pre-August 1980 and is outside the searchable database. Its absence from Assignment Center is a coverage artifact, not evidence of no transfer.
2. The patent expired 1993-12-07 (17-year term from the 1976-12-07 issue date; Google Patents records "Anticipated expiration 1993-12-07," legal status Expired – Lifetime). It therefore had no remaining enforceable term during the entire modern NPE era. This is dispositive for most of the troll-pattern signals below.

Identifier-collision warning (carried forward from the prior section). The numeral string "3996429" is not unique in public registries. It appears as a European patent validation number (11) 3996429 for an LG Electronics application in the Belgian Recueil des brevets (https://economie.fgov.be/sites/default/files/Files/Intellectual-property/Recueil%20brevets/recueil-brevets-2025-07.pdf), and it has appeared in a Lithuanian legal-notice serial. Any grep-style search of assignment or ownership registries for "3996429" must be manually de-duplicated, or it will produce false assignee hits that have nothing to do with US 3,996,429. I found no such false hits masquerading as US assignment records, but the collision is real and worth stating.

Inventors

Notes on pattern. Both inventors are the same pair named on US 4,029,916 (Northern Electric Company Limited, priority 1975-04-18, issued 1977-06-14) — the sibling filing on the same priority date, already flagged in the prior summary as the closest same-family relative. A single-assignment filing by two co-inventors to their corporate employer is the ordinary 1970s Canadian industrial-laboratory pattern (Northern Electric / BNR, Montreal–Ottawa corridor).

I have no evidence either way on inventor retention or departure within 12 months of filing. I did not find personnel records or SEC-style disclosures covering Chu or Reedyk, and I explicitly decline to infer a fire-sale precursor from silence. Not determinable.

Original assignee

Northern Electric Company Limited (Canada) — named on the issued patent front page.

• Primary line of business: Canadian manufacturer of telephone and telecommunications equipment; by the 1970s a major switching-system and terminal-equipment supplier and parent-lineage entity of what became Northern Telecom and then Nortel. Source: https://en.wikipedia.org/wiki/Nortel (the company opened its first US factory in Michigan in 1972 and was shipping early digital switching systems in this era).
• Did they ship a product embodying the claims? Effectively yes, on the record. The specification is written as a production-oriented design — stamped stainless-steel contact plate, printed-circuit board base, spacer, cover, pushbutton, and a ten-button "2 + 5" telephone dial-pad layout with a documented binary code table — i.e., a design for a telephone set dial pad rather than a laboratory curiosity. The citation history confirms commercial uptake: the forward citations include US 4,486,637 (Northern Telecom Limited) and the assignee family was still filing in this space into the 1980s. I could not, however, tie the claims to a specific model number, so treat "shipped a product" as strongly supported by context but not documentary from a primary catalogue record.
• Current status: Northern Electric was renamed Northern Telecom Limited in March 1976. The successor lineage — Nortel Networks — filed for bankruptcy protection in January 2009 and its patent portfolio was sold at auction in 2011. The original assignee entity no longer exists in any operating form. Sources: https://en.wikipedia.org/wiki/Nortel (name change); Google Patents front page for successor-assignee lineage.

Assignment timeline

No assignment records were retrieved for US 3,996,429 from the USPTO Assignment Center. Stating this plainly, as instructed.

What I can say with precision about why, and what the record limits are:

• The Assignment Center database begins August 1980. This patent's realistic assignment events (1975 inventor-to-corporate assignment; 1976 change of name Northern Electric → Northern Telecom Limited) both predate the searchable window. They would be indexed, if at all, only in the separate pre-1980 historical database requiring Assignment Recordation Branch or NARA contact — not reachable through the web interface.
• No reel/frame numbers are reported here because I did not find any. I will not synthesize plausible-looking reel/frame citations; per the operating rules, a fabricated reel/frame is worse than an acknowledged gap.
• The only tangible post-issuance ownership datum I have is the Google Patents "Current Assignee: Nortel Networks Ltd" field on https://patents.google.com/patent/US3996429/en. Google expressly disclaims that list ("The listed assignees may be inaccurate... has not performed an analysis"). Read alongside the Google Legal Events entries for this patent (Application filed 1975-04-18 → Application granted / Publication 1976-12-07 → Anticipated expiration 1993-12-07, status Expired – Lifetime), the most parsimonious reading is that the Nortel attribution reflects corporate name-change lineage, not a recorded sale. A "Change of Name" conveyance of this kind would typically be a single blanket recording covering thousands of patents by reel/frame — which I could not locate for this patent in any indexed source.
• Correspondent of record: none found. There is consequently no repeat-player attorney signal to mine, which is the primary analytical loss from the access gap.

Per the task instruction I would normally stop here. I am continuing with the remaining sections because (a) my null result is a retrieval gap rather than a confirmed empty record set, and (b) the signal analysis below is driven by an independent, verifiable fact — the 1993-12-07 expiry — that does not depend on reel/frame data. I have marked every cell that rests on the access gap rather than on evidence.

Timeline diagram

timeline
    title Ownership of US 3996429
    1975 : Application filed by two inventors
         : Rights assigned to Northern Electric
    1976 : Patent issued on 7 December
         : Northern Electric renamed Northern Telecom
    1993 : Patent term expires on 7 December
    2009 : Nortel files for bankruptcy protection
    2011 : Nortel portfolio auctioned to Rockstar
    2026 : No assignment records found online

Note: the 1975 assignment and the 1976 name change are the expected corporate events for a patent of this vintage, not reel/frame-verified records. The 2009 and 2011 events are Nortel-lineage corporate context; because this patent expired in 1993, it was not an enforceable asset in the 2011 auction (see signal 6).

NPE / troll-pattern signals

Suppressed signals worth naming. Several classic NPE tells simply cannot fire here, and the reason matters: (a) a patent with no enforceable term since 1993 generates no litigation, so there is no venue-driven transfer for signal 5; (b) with no post-1980 recordings retrieved, the highest-value tell in the entire playbook — the recurrent recording attorney — is unavailable; (c) there is no Continuation-in-Part, continuation, or divisional chain to trace, because the sibling US 4,029,916 issued independently in 1977 and I found no continuation filings from either. This is a small, self-contained 1975-era two-patent family, not a portfolio engineered for assertion.

Verdict

Insufficient data.

The chain terminates at the original assignee's corporate lineage (Northern Electric Company Limited → Northern Telecom → Nortel Networks), and I retrieved no Assignment Center records at all — a null driven by the fact that the USPTO's online assignment database only reaches back to August 1980, which post-dates this patent's only likely recorded transfer. There is no shell entity, no known asserter, no chained-LLC cascade, and no pre-litigation transfer to cite. Independently dispositive: the patent expired 1993-12-07, meaning it was never assertable during the NPE era and could not have been an enforceable asset in the 2009–2011 Nortel bankruptcy sale. The correct label is therefore insufficient data — and on the balance of what is known, the honest secondary characterization is an operating-company ownership history that simply ran out of term before the modern assertion market existed, not an NPE chain.

Verification links. USPTO Assignment Center: https://assignmentcenter.uspto.gov/ · legacy interface: https://assignment.uspto.gov/patent/index.html (search by patent number 3996429). For pre-August 1980 records that the web interface cannot surface, contact the USPTO Assignment Recordation Branch (571-272-3350) and, for records recorded before 1957, NARA at 4205 Suitland Road, Suitland, MD 20746.

Confidence statement. I have high confidence in the coverage-window finding (August 1980 database start) and the expiry date (1993-12-07), both of which are well-sourced. I have low confidence in the affirmative proposition "zero assignments were ever recorded," and I have no confidence whatsoever in any specific reel/frame, correspondent name, or filing date for this patent — none were found, and none are asserted here.

I'll search for the patent and each of its cited prior-art references.

Now the remaining cited references.

Search result — direct hit confirmed

The search returned the target document itself, not a similar number:

• US 3,996,429 A — Multi-contact push-button switch having plural prestressed contact members designed to provide plural circuit simultaneous switching inputs
• Full text: https://patents.google.com/patent/US3996429/en · Grant PDF: https://patentimages.storage.googleapis.com/96/82/ba/fcd37e4a710c17/US3996429.pdf · Third-party mirror: https://uspto.report/patent/grant/[3996429](/patent/3996429)

All identifiers resolve literally as written: application US 05/569,359, filed and priority 1975‑04‑18, granted 1976‑12‑07, inventors Pak‑Jong Chu and Cornelis Wilfred Reedyk, original assignee Northern Electric Company Limited. No similar-number substitution occurred.

Framing the §102 question (important)

Claim 1 is the only independent claim, and it is a combination claim. Under 35 U.S.C. §102 (pre‑AIA, which governs a 1975 filing), a reference anticipates only if it discloses every element as arranged in the claim. The four load-bearing elements of claim 1 are:

1. A flat spring plate with plural integral switch positions, each a prestressed domed portion completely surrounded by, and circumferentially constrained by, a flat portion;
2. Radial webs with an upwardly bent portion at the junction with the surrounding flat plate, bending inward of that bend to produce the return force;
3. Cantilevered spring contact members extending from the center portion, one between each pair of webs, reaching a shorter radial distance than the webs, with their contact areas positioned below the central contact;
4. A circuit member below with peripheral contact areas plus a further central contact area, so that the peripheral (code) contacts close before the central contact.

I note up front that no single cited reference contains all four. My per-reference verdicts below reflect that, and I mark explicitly where a reference is really §103 material.

Citation-table discrepancy to flag: the front page lists the citations twice — a "Citations (11)" table and a "Patent Citations (12)" table. The 11-entry version omits US 3,742,157; the 12-entry version includes it. My analysis uses the 12-entry list (the fuller one). Also note Google Patents' "publication date" column is the US grant date, not a pre-grant publication date (pre‑1975 US applications generally had none).

The 12 cited references

Statutory status: Nos. 1–5, 7, 8, 10, 11 were granted before 1975‑04‑18 → classic §102(a)/(b) art. Nos. 6 and 12 were granted after it, but were filed before it (1971‑08‑23 and 1974‑12‑09) → §102(e) "secret" prior art only. US 4,029,916 does not appear in this list — it sits in the Cited By table with the same 1975‑04‑18 priority and assignee Northern Electric, so it is a related family document, not prior art, and cannot anticipate.

Per-reference analysis

1. US 3,133,170 — Nanninga, "Snap switch having a concavo-convex diaphragm"

Description (verified from the grant PDF and RPX record): An annular enclosure with an insulating base carrying a centrally disposed contact element 22 and peripheral terminal assemblies 24, 26. A resilient concavo‑convex diaphragm 34 is seated in a V‑groove retainer ring 32. The diaphragm has a central area 38 carrying a movable contact 36 and an outer/rim area separated by arcuate slots 40 or slits 78 joined by tabs 33. An actuator 46 with a radially extending body and peripheral subtending legs 44 engages the peripheral zone of the central area to produce snap action of the central area and rapid engagement/disengagement with the fixed central contact. Crucially, the specification states the diaphragm "may be of annealed, unstressed material" and that "there is no requirement for a pre-stressed condition."
§102 verdict: Anticipates none of claims 1–3. It lacks the plural cantilevered spring contact members between radial webs (element 3), the circuit member with plural peripheral contact areas (element 4), the plural-pushbutton integral plate, and the sequential closure order — and its express teaching away from pre-stressing cuts against claim 1's "prestressed" limitation. Properly, it is §103 art for the snap-action dome + central movable contact concept. URL: https://insight.rpxcorp.com/patent/[US3133170A](/patent/US3133170A)

2. US 3,600,528 — Lematex Corp. (Leposavic), "Multiple switch construction"

Description (verified from the grant PDF / Google Patents): Multiple switches as a flat laminated panel: a non‑conductive base panel 20 with a photographically printed/etched circuit forming fixed contact elements 21, a screen of spacers 23 dividing separate switch areas 22, and a resilient cover 26 whose underside carries a conductive film 28 acting as the movable contact. Finger pressure at an area momentarily closes the switch; the cover snaps back on release. Later figures show ring‑type and center‑type flexible contact elements, and a gang stamping 35.
§102 verdict: Anticipates none of claims 1–3. It discloses plural switch positions on a common panel and a circuit member below with spacers — but the cover itself is the actuator (no separate pushbuttons, no cover apertures, no interengaging retaining formations as in claim 2), the movable contact is a conductive film on a membrane rather than a prestressed domed plate with radial webs and inter‑web cantilevered contacts, and there is no central series/delayed contact. §103 art for elements 1 and 4 broadly. URL: https://patents.google.com/patent/US3600528

3. US 3,742,157 — Lematex Inc. (Leposavic), "Keyboard switch assembly with improved movable contact"

Description (verified): A same‑family improvement over '528. Movable switch contact element with a moment arm imparting a rotary wiping action on depression; a retainer frame 34 registering with the spacer screen; a separate spring member 50a/58 controlling pre‑travel, pre‑load and over‑travel; push‑button return springs; air pockets to suppress bounce. The contacts 39a move downwardly with rotary wiping action until contact is made with the fixed contact elements 22a.
§102 verdict: Anticipates none of claims 1–3. Notably relevant because US 3,996,429's own specification recites "radial wiping action by the contacts on the contact areas," and '157 is the art that already shows wiping; but '157 has no prestressed dome/radial webs and no sequentially closed central contact. §103 art. URL: https://patents.google.com/patent/[US3742157A](/patent/US3742157A)/en

4. US 3,761,944 — Alps Electric (Shimojo), "Binary code generator"

Description (verified from the grant PDF / Google Patents): A multi‑row pushbutton encoder. A circuit board carries a plurality of conductive leads, one per binary bit, each wired to an output terminal; interlaced lead groupings (one per decimal digit) are laid out with substantial point symmetry, "each individual lead traversing a plurality of radial paths spaced around the center of the grouping." A resilient keyboard sheet carries a conductive rubber contact that, on key depression, bridges all leads of that grouping to a voltage source, producing a selected 4‑bit code. The applicants expressly describe this as eliminating diode matrices.
§102 verdict: Anticipates none of claims 1–3, but this is the single most relevant reference on the binary‑coding function of claim 3 (predetermined contact configuration → predetermined binary code on depression, with radially arranged lead branches). It fails claim 3 because claim 3 carries all of claim 1's and claim 2's limitations, and '944 uses one bridging pellet per key, not plural inter‑web cantilevered spring contacts with a delayed central gating contact, and has no prestressed dome. Strong §103 art against claim 3's encoding aspect. URL: https://patents.google.com/patent/US3761944

5. US 3,699,294 — Flex Key Corp., "Keyboard, digital coding, switch for digital logic, and low power detector switches"

Description: The title and front‑page data are from the 3,996,429 citation table; I could not retrieve this reference's own text before my search budget was exhausted, so the description below is background knowledge, lower confidence. It concerns a flexible keyboard with digital coding of key closures for logic circuits and low‑power detector switching.
§102 verdict: No anticipation of claims 1–3. Even crediting a digital‑coding keyboard, it does not disclose the prestressed domed spring plate with radial webs, the inter‑web cantilevered spring contacts whose contact areas sit below the central contact, or the delayed central/common contact. §103 art at most, and chiefly against claim 3's general "binary coding by key" idea.

6. US 3,941,953 — Hewlett‑Packard, "Keyboard having switches with tactile feedback"

Description: Title/date verified from the citation table; body text not retrieved → lower‑confidence description. As titled, it concerns keyboards whose keys give tactile (snap/feel) feedback.
§102 verdict: No anticipation. Relevant to the "positive feel"/snap‑action object recited in 3,996,429's background, not to the claimed structure. Statutory note: this reference issued 1976‑03‑02, after the 1975‑04‑18 filing, so it can only be §102(e) art via its 1971‑08‑23 filing date. §103 art only.

7. US 3,826,882 — Olivetti & C. SpA, "Electric keyboards for office machines"

Description: Verified title/date from the citation table; body not retrieved → lower‑confidence description. Concerned with keyboards for office machines (typewriters/calculators).
§102 verdict: No anticipation of claims 1–3. §103 art for general keyboard-array context only.

8. US 3,898,421 — Matsushita Electric Industrial, "Push button switch with elastic conductive sheet"

Description: Verified title/date; description lower confidence (body not retrieved). Uses an elastic (conductive) sheet as the movable bridging contact — the same class of element that 3,996,429's own FIG. 8 description acknowledges ("Contact 48 can be a conductive rubber pellet, conductive epoxy resin, or other means").
§102 verdict: No anticipation. A single bridging sheet is not the claimed prestressed dome + inter‑web cantilevered spring contacts, and it provides no delayed central contact. §103 art.

9. US 3,870,840 — Sits Soc. Italiana Telecomunicazioni Siemens (Rivetta et al.), "Switch pushbutton‑type digit keyboard switch with leaf spring contact and actuator biasing means on common conductive frame"

Description: Verified title/date from the citation tables (this reference also appears as "3870840 April 1975 Rivetta et al." in the reference list of US 4,146,767). It concerns a digit keyboard using leaf‑spring contacts on a common conductive frame with actuator biasing means.
§102 verdict: No anticipation of claim 1, but this is meaningful §103 art for the "common flat spring plate / common frame carrying plural switch positions," the leaf‑spring contact members, and the return‑bias ("return force") element of claim 1. It does not show the prestressed dome bounded by and circumferentially constrained by a flat portion, nor the delayed central gating contact.

10. US 3,800,104 — Becton, Dickinson & Co., "Low profile keyboard switch assembly with snap action cantilever contact"

Description: Verified title/date. Squarely directed to a snap‑action cantilever contact in a low‑profile keyboard.
§102 verdict: No anticipation of claim 1, but high‑relevance §103 art for the "cantilevered … contact members … with snap action" element. It lacks the prestressed domed plate, the radial webs with upwardly‑bent junctions generating return force, and the sequential central contact.

11. US 3,796,843 — Bomar Instrument Corp., "Calculator keyboard switch with disc spring contact and printed circuit board"

Description: Verified title/date. Combines a disc (dome) spring contact with a printed circuit board — the two structural pillars of claim 1.
§102 verdict: No anticipation of claim 1. An essential difference remains: the claim requires plural cantilevered spring contact members arranged one between each pair of radial webs, with their contact areas below the central contact so that peripheral closure precedes central closure. '843 teaches a disc spring making a single per‑key connection, without the two‑stage delayed‑common‑contact gating. §103 art.

12. US 3,941,964 — Bowmar Instrument Corp., "Push‑button type binary switch device"

Description: Verified title/date. A push‑button binary switch — relevant to both the pushbutton/switch‑position aspect of claim 1 and the binary‑code aspect of claim 3. Statutory note: filed 1974‑12‑09, granted 1976‑03‑02 — granted after 3,996,429's filing date but filed before it, so it is §102(e) art only.
§102 verdict: No anticipation of claims 1–3. Like '944, it addresses binary output but not the delayed central common contact or the dome/webs/cantilevered‑contact geometry. §103 art; possibly combined with '944 against the binary‑code limitation of claim 3.

Overall §102 conclusion

Claim 1 is not anticipated by any of the twelve cited references. No reference discloses the combination of (i) a prestressed domed portion completely surrounded by and circumferentially constrained by the flat portion of an integral flat spring plate, (ii) radial webs whose upwardly bent junction portions define the point of flexure generating the return force, (iii) cantilevered spring contact members, one between each pair of webs, having contact areas below the central contact, and (iv) the resulting two‑stage closure in which all peripheral code contacts pre‑close and are then gated by the delayed central common contact.

Because claims 2 and 3 depend from claim 1 and incorporate all of its limitations, they likewise are not anticipated. The cited art is best characterized as a §103 evidentiary cluster:

• Encoding function (claim 3): US 3,761,944 (Alps) and US 3,941,964 (Bowmar) — strongest.
• Dome/disc spring + printed circuit board (claim 1 structural core): US 3,796,843 (Bomar), and US 3,133,170 (Nanninga) for the snap‑action concavo‑convex diaphragm.
• Cantilever/snap‑action contact: US 3,800,104 (Becton Dickinson).
• Common frame + leaf springs + return bias: US 3,870,840 (Rivetta/Siemens).
• Membrane/laminated multi‑switch panel + circuit board + spacer + cover: US 3,600,528 and US 3,742,157 (Lematex).
• Tactile feedback: US 3,941,953 (H‑P).
• Elastic conductive sheet bridging: US 3,898,421 (Matsushita).

What distinguishes 3,996,429 from all of these — and what the examiner evidently relied on — is the series delayed central contact using a common contact in series with the encoding contacts, eliminating the time‑delayed strobe/encoding logic. That is visible in the forward citation treatment of the patent: EP 0 084 734 A1 / US 4,400,596 and EP 1 087 416 A1 (Molex) each cite US 3,996,429 as an [XA] reference for "sequentially closeable/closable contacts." URLs: http://data.epo.org/gpi/EP0084734A1 and http://data.epo.org/gpi/EP1087416A1

Confidence and caveats

• Verified by search text: US 3,133,170, US 3,600,528, US 3,742,157, US 3,761,944 (bodies and/or claims retrieved), plus the target patent's own record and the two EP forward‑citing documents.
• Lower confidence — descriptions for US 3,699,294, US 3,941,953, US 3,826,882, US 3,898,421, US 3,870,840, US 3,800,104, US 3,796,843 and US 3,941,964 are drawn from their titles and front‑page data in the 3,996,429 citation table plus general background knowledge; I did not retrieve their full texts before reaching my search limit. Titles, dates and assignees are reliable as transcribed; the one‑line technical descriptions for those eight are not independently verified and should be confirmed against the documents.
• §102 vs §103: My "no anticipation" verdicts reflect the strict all‑elements rule. A §103 obviousness analysis would likely combine e.g. Nanninga/'843 (dome + board) with '104 (cantilever snap contact) and '944/'964 (binary encoding); I have not performed a full motivation‑to‑combine analysis here.
• Dates: All filing/priority and grant dates are as listed by Google Patents and are consistent with the numbers cited; I have not independently confirmed them against the USPTO PatentCenter/Global Dossier records.
• I reiterate from the prior section: I found no 2026 CAFC, IPR/PGR or district‑court record naming US 3,996,429, which is expected for a patent that expired 1993‑12‑07.

I'll ground the prior-art references before building the combination analysis.

Strong hit: US 3,941,964 (Bowmar) appears to disclose the delayed central contact. Let me verify the remaining combination members.

Obviousness Analysis — US 3,996,429 under 35 U.S.C. § 103

Builds on prior sections (bibliographic data, claim overview, prior-art list). This section does not repeat the claim summaries; it treats them as given and maps them to references.

Governing law applied (pre-AIA, since the application was filed 1975-04-18, before the AIA's 2013 first-inventor-to-file change): Graham v. John Deere Co., 383 U.S. 1 (1966) (scope/content of prior art, differences, PHOSITA level, secondary considerations); KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) (motivation may be supplied by market demand, design incentives, "obvious to try," and predictability); MPEP § 2143 (rationales A–G and the TSM alternative).

1. Prior-art status of each reference (pre-AIA § 102 flavor)

All twelve references on the face of the patent predate the 1975-04-18 filing. The two that matter most:

⚠️ Important negative — do NOT use US 4,029,916 as § 103 prior art. The sister case flagged in the prior sections (Northern Electric, Multi-contact push-button switch and plural embodiment for keyboard switch assembly) carries the same 1975-04-18 priority date as US 3,996,429. A same-day-filed document is not "filed … before the invention by the applicant" under pre-AIA § 102(e), and common ownership would independently disqualify it from § 103 via pre-AIA § 103(c). Any obviousness theory built on US 4,029,916 is void ab initio. It is useful only as evidence of the inventors' own contemporaneous understanding of the art.

2. What the cited art actually discloses (verified against full text)

US 3,941,964 — the most dangerous reference. This teaches the entire logical core of the invention:

• An "overcenter, snap-acting, 'oilcan' diaphragm switch element 12… movable between a first inactive position and a second overcenter operative position."
• A center contact dimple 14 in registry with a common terminal pad 28.
• Relatively short, radially arrayed feet 36 in registry with terminal pads 38.
• The critical delayed-contact teaching: "diaphragm element 12 and leg members 16 are proportioned and arranged so that diaphragm element 12 passes through a third, intermediate position in which feet 36 engage terminal pads 38 prior to engagement of center contact dimple 14 with common terminal pad 28." The reference then states it is "important that the center contact dimple 14 make contact last with feet 36 making contact prior to that time, and that the center contact dimple 14 be the first to break contact upon release."
• Binary coding by selective connection: "external conductor 46 may be connected to terminal pad 38 and external conductor 44 may be connected to terminal pad 26 to provide a binary coded signal corresponding to the digital numeral three" … "in accordance with an 8421 binary code … for the digits zero through sixteen."
• Explicit keypad application: "a plurality of such binary switches may be assembled in a keyboard resembling a calculator keyboard or push-button telephone dial."

That is, in substance: snap dome + peripheral code-select contacts + delayed common contact in series + 8421 binary output on a single-sided PCB + telephone-dial keypad. This is the same problem solved by the same mechanism for the same application.

US 3,796,843 (Bowmar). Discloses the structure of the movable member: a conductive snap-acting dome 14 with downward projections/dimples 64 at apices 60 (peripheral contact areas) and a center projection/dimple 66 in registry with a center terminal 32; a single-sided printed circuit board; a retaining sheet with openings and a cover member 20 with openings 46 receiving push-buttons 22; "Application of manual finger pressure on a particular push-button 22 results in the application of force through cover sheet 18 on a respective snap-acting dome switch member 14 thereby actuating the switch member to its non-overcenter deflected position." See also its reissue RE30923 and RE29440.

US 3,742,157 and US 3,600,528 (Lematex, Leposavic). Teach the flat stamped spring plate carrying multiple integral switch positions ("movable contact elements are preferably formed in a gang stamping 32 … on a sheet of tempered beryllium copper"; claim 5 of US 3,600,528: "a flexible electrically conductive element comprising a stamping of substantially flat-rigid electrically conductive material having frame-like portions … and integral movable switch elements suspended from said frame-like portions"). They also teach spider-like radial fingers engaging ring-shaped fixed contacts with a rotary wiping action, and — critically — sequenced contact levels: "it is desirable to provide one fixed contact 23 in any given switch area 26 at a level slightly below the level of the remaining contacts. As the movable contacts are moved toward and away from the fixed contacts, the lower fixed contact is the last to make and the first to break the circuit and may be utilized in computer input devices as a control function."

US 3,800,104 (Becton, Dickinson). Teaches a "thin resilient sheet that is elastically stressed into a three-dimensional configuration" forming cantilever spring members with a prestressed portion that snaps between two curvatures, made of "phosphor bronze, beryllium copper or stainless steel," giving both audible/tactile feedback and — expressly — "Switch contacts are only engaged or disengaged after the switch has snapped into a stable configuration."

US 3,133,170 (Nanninga) — concavo-convex diaphragm snap switch (the basic monostable dome). US 3,941,953 (Hewlett-Packard) — keyboard with tactile-feedback switches. US 3,761,944 (Alps) — "Binary code generator." US 3,699,294 (Flex Key) — "Keyboard, digital coding, switch for digital logic, and low power detector switches." US 3,828,882 (Olivetti) — electric keyboards. US 3,898,421 (Matsushita) — push-button switch with elastic conductive sheet. US 3,870,840 (Sits/Siemens) — pushbutton digit keyboard with leaf spring contact and actuator biasing.

Verification caveat: I retrieved full text for US 3,941,964; 3,796,843 (via RE30923); 3,742,157; 3,600,528; and 3,800,104 (via US 3,786,205 cross-usage). For US 3,761,944, 3,699,294, 3,133,170, 3,941,953, 3,828,882, 3,898,421, 3,870,840 I am relying on the titles as published on the patent page plus general field knowledge; my search budget was exhausted before full-text retrieval. Treat the descriptive attribution for those six as lower-confidence.

3. Claim 1 — limitation mapping against the primary combination

Primary combination: US 3,941,964 + US 3,796,843 + US 3,742,157 (+ optionally US 3,800,104).

Every limitation of claim 1 is disclosed, and the single most distinctive limitation — the delayed common/central contact — is disclosed in terms, for the identical purpose, in US 3,941,964. The remaining differences are (i) forcing the dome to be monostable via the upwardly bent junction portion 21, and (ii) the specific radial-web/cantilever-finger geometry of FIG. 1. Both are addressed below.

4. Concrete § 103 combinations

Combination A (strongest): 3,941,964 + 3,796,843 + 3,742,157

Renders claim 1 obvious; renders claim 3 obvious alone (see §6).

Motivation to combine — the express-teaching rationale. US 3,941,964 itself cites and builds on US 3,796,843: "U.S. Pat. No. 3,796,843 assigned to the Assignee of the present application discloses a keyboard switch assembly employing an overcenter, snap-acting diaphragm switch element and a single-sided printed circuit board." This is not an after-the-fact reconstruction of a motivation; it is the reference's own stated starting point. KSR treats an express identification of a reference in the art as a strong motivation, and same-assignee/same-field proximity strengthens it further (both are Bowmar Instrument calculator-keyboard patents).

Motivation to add Lematex '157/'528. The Lematex gang-stamping teaches how to take the single-dome, single-switch Bowmar structures and multiply them into a single stamped spring plate serving many key positions — precisely the "plurality of integral switch positions" limitation. It also supplies the radial-finger-with-wiping-contact architecture and, independently, the "lower fixed contact is the last to make and the first to break" teaching that corroborates the delayed-contact principle.

Combination B: 3,133,170 + 3,600,528 + 3,941,964 + 3,800,104

Substitutes Nanninga's concavo-convex snap diaphragm as the primary dome with Lematex's multi-position stamped plate and Bowmar's delayed common contact, and uses Becton Dickinson '104 for the prestressed stainless-steel spring sheet and contact-after-snap operation.

Combination C: 3,761,944 (Alps, "Binary code generator") or 3,699,294 (Flex Key, "digital coding … for digital logic") + 3,796,843 + 3,941,964

Directed particularly at claim 3; supplies independent evidence that outputting binary codes directly from a keypad's contact pattern was a known and pursued objective in the same 1971–1974 window, independent of Bowmar.

5. Why a PHOSITA would have combined these (KSR rationales, MPEP § 2143)

1. Rationale A — combining known elements with predictable results. Each element (snap dome; PCB with peripheral + central pads; stamped multi-position spring plate; selective pad connection for binary code; pushbutton cover) performs its known function; the assembly is the predictable sum.
2. Rationale B — simple substitution. Bowmar '843's triangular dimpled dome and Bowmar '964's overcenter oilcan diaphragm are interchangeable snap elements in the same keyboard; the specification of '429 itself treats the monostable/bistable character as a design/clamping variable rather than an invention: "Clamping pressure … affect the return force and can alter a bistable contact plate to a monostable contact plate, when clamped."
3. Rationale C — known technique, same improvement. Lematex had already solved multi-key stamping; Bowmar had already solved sequenced contacts.
4. Rationale D — a known device ready for improvement. '843 is literally identified by '964 as the starting point to be improved.
5. Rationale F — design incentives / market forces. The '429 background states the constant demand: "switches be made more compact, requiring a minimal space in the apparatus. At the same time the switch should be simple and reliable." Calculator and telephone keypad cost/size pressure in the early 1970s is exactly the KSR "design incentive" category.
6. Rationale G — obvious to try. The problem of non-simultaneous contact make was recognized in the art (Lematex '157's lower-level contact; '429's own admission that "manufacturing tolerances making it impossible to align the contact fingers 12 to 15 such that they all make contact simultaneously"). Two solutions were known — stagger the pad heights, or delay a common contact. Substituting one for the other is a finite number of identified, predictable solutions.
7. Reasonable expectation of success. Both references are complete, operative, commercially deployed keyboard switches; no unpredictable technology is implicated.

Argument-killer note for claim 1's "monostable return force": Bowmar '964 already describes snap-over and snap-back — the structure retains a return tendency. Even absent the '21 upward-bend geometry, one of ordinary skill would select a monostable dome because a keyboard requires momentary (not latching) operation, and Lematex '157 supplies spring stampings "providing a positive spring return for each of said depressible means."

6. Dependent claims

Claim 2 — cover with apertures + interengaging retaining formations

Would be obvious over Combination A in view of US 3,796,843 (cover member 20 with openings 46 receiving push-buttons 22; retaining sheet 16 with openings 42 locating the dome) and US 3,742,157 (claim 4: "depressible means is an array of push buttons supported for linear movement in a guide plate"), or US 3,800,104 (keys retained in a frame member). The '429 specification's own words confirm this is routine packaging: the flange 36 / recess 33 arrangement is described as conventional structure with no asserted surprise. Preserving buttons in a keyboard cover is a known, purely mechanical problem with a known solution — Rationale A/B.

Claim 3 — binary coded signal system with predetermined contact configuration

Would be obvious over Combination A and, independently, over US 3,941,964 alone:

• '964 discloses binary coded signal output from a snap-dome switch ("external conductors … connected to preselected ones of the remaining terminal pads").
• It expressly frames the switch as usable in "a keyboard resembling a calculator keyboard or push-button telephone dial."
• The selective-connection-of-contact-areas concept is therefore old; the '429 patent's own contribution is described as eliminating interface logic ("encoding circuits are eliminated"), i.e., an economy, not an unobvious technical step.
• US 3,761,944 and US 3,699,294 provide corroborating art on direct binary encoding from keypad contacts.

Because the printed-circuit-pattern-selection step is a packaging/routing choice implemented "using standard printed circuit board layout techniques" (the specification's own words), the claim 3 narrowing is squarely a predictable, art-recognized implementation detail.

7. Weaknesses in the obviousness case — must be assessed honestly

(a) The examiner considered the two best references and still allowed the claims. US 3,941,964 (Bowmar) and US 3,796,843 (Bowmar), plus US 3,133,170, US 3,600,528, US 3,742,157, US 3,699,294, US 3,761,944, US 3,941,953, US 3,828,882, US 3,898,421, US 3,870,840, US 3,800,104 — all twelve appear in the "Patent Citations" list of US 3,996,429. Any ex parte or inter partes challenge therefore carries no novelty-of-reference advantage; the case must be a new combination or a new motivation argument, and the patentee will argue that the examiner had exactly these pieces and found the whole non-obvious.

(b) Potential teaching-away / criticality argument for the '21 junction geometry. The '429 specification goes out of its way to explain that the single stable state is "created by the particular formation of the upwardly bent junction portion 21" and that it "provides a positive return force on the central portion at all times." A patentee will argue this is a specific structural expedient with asserted functional consequence (guaranteed momentary operation), not a mere design choice. The counter is the specification's own admission that clamping the flat portion 11 can convert a bistable plate to a monostable one, and that "Other formations for ensuring a positive return force at all times can be provided" — i.e., the patentee concedes the result is attainable by multiple known routes.

(c) Structural integration argument. Claim 1 requires the domed portions and cantilevered contact members to be integral with one flat spring plate ("surrounded by a flat portion of said flat spring plate," "integral therewith"). Bowmar '964 discloses a single diaphragm; the Lematex references supply the gang-stamped multi-position plate. The patentee may argue that no reference suggests co-locating many monostable delayed-contact domes on one stamped sheet. Lematex '528 claim 5 and '157's "gang stamping" language cut strongly against this, but it is the patentee's best structural argument.

(d) No verified secondary considerations. I found no evidence of long-felt need, failure of others, licensing, or commercial success attributable to these claims, and the patent expired 1993-12-07. Nexus-free commercial-success arguments would be weak — but a challenger also cannot point to negative secondary considerations. This factor is currently neutral, not favorable.

(e) Date discrepancy in the task framing (flagged per instructions). The patent-page header supplied in this analysis states "Today's date: 2026-09-30," while the task statement says "Current Date: April 26, 2026." This does not affect any analysis above (all relevant dates are 1971–1976 and 1993), but I note it rather than silently reconciling it.

8. Conclusion

• Claim 1 is highly vulnerable to § 103 over US 3,941,964 + US 3,796,843 + US 3,742,157, with US 3,800,104 as an optional fourth reference. The delayed-common-contact feature — the asserted heart of the invention — is disclosed in express terms, for the same purpose, in the same field, by US 3,941,964, which itself identifies US 3,796,843 as its starting point.
• Claim 2 is vulnerable over the same combination plus US 3,796,843 / US 3,742,157 / US 3,800,104 for the cover-and-button-retention expedient.
• Claim 3 is vulnerable over the same combination and independently over US 3,941,964 (which discloses 8421 binary coded output for a telephone-dial-like keyboard), with US 3,761,944 and US 3,699,294 as corroboration.
• The single largest practical obstacle is not the art — it is that the examiner cited every one of these references and allowed anyway. A realistic challenge must therefore be framed as a new combination with an articulated KSR motivation (the same-assignee/express-incorporation link between the two Bowmar references is the strongest available motivation), not as a re-presentation of individual references.

9. Statement of uncertainty

• Full text was verified only for US 3,941,964, US 3,796,843/RE30923, US 3,742,157, US 3,600,528, and US 3,800,104 (the last partly via its sibling US 3,786,205). Content attributions for US 3,761,944, US 3,699,294, US 3,133,170, US 3,941,953, US 3,828,882, US 3,898,421, US 3,870,840 rest on their published titles and field knowledge; that part of the mapping is not fully verified.
• I did not locate any litigation, IPR/PGR, or reexamination record for US 3,996,429, consistent with its 1993 expiry; the absence is not proof that none exists.
• I have not independently verified the pre-AIA § 102(e) "different inventive entity" premise for US 3,941,964 beyond differing named inventors and assignee (Bowmar vs. Northern Electric) — a nuance worth confirming from the USPTO assignment records before relying on it.
• All quoted claim language is taken verbatim from the claims section of the supplied patent text; all quoted reference language is from the URLs cited above.