US 9331488

I'll search for authoritative information on this patent, including litigation/docket data.

Let me check for Federal Circuit (CAFC) activity and any PTAB/IPR challenges specific to this patent.

Let me check for PTAB/IPR challenges and any Federal Circuit appeals.

Method note

I do not have direct PATCTS/PACER/PatentCenter API access here; I ran live web searches (Google Patents, Justia, RPX Insight, Patexia, CourtListener, PACER Monitor, Ex Parte AI case pages, Ramey LLP case announcements, IP-NAVI). Everything below about the patent document is authoritative from the Google Patents full text you supplied plus corroborating mirrors. Everything about litigation is from the docket/aggregator sources cited and should be treated as secondary evidence.

⚠️ Date discrepancy flagged: your prompt says the current date is April 26, 2026, but my system date is September 19, 2026, and some search results carry timestamps after April 26, 2026 (e.g., an AO-120 report entry dated 2026-08-26 in the Home Depot docket, and a Generac complaint dated 08/27/26). I report those items but they may post-date the April 26, 2026 snapshot you intended.

US 9,331,488 B2 — Bibliographic data

Abstract (as issued): An enclosure design is disclosed to accommodate and support the unique features and capabilities of the Smart and Scalable Power Inverters or Mini-Inverters that have multiple input channels to easily connect to multiple solar PV panels, invert the DC power to AC power, and daisy chain together to generate AC power to feed the power grid or supply power to electrical devices. Further disclosed is a message system using LEDs mounted on the enclosure to indicate the system status and the status of each input channel.

Specification highlights (grounding the claim analysis): FIGs. 1–3 (4-channel and 12-channel enclosure designs), FIG. 4 (grid-tied single-phase block diagram), FIG. 5 (off-grid "AC Master" three-phase block diagram), FIGs. 6–8 (main microcontroller program and LED status subroutines). The LED message system is defined by Tables 1–4 (Cases 1–10 for grid-tied; Cases 11–20 for off-grid AC Master): solid green = working; flashing green = working with warnings; solid red = system error; flashing red = grid/AC-load error; off = not enough PV power / waking up. The patent cross-references Ser. No. 12/837,162 and Provisionals 61/442,991 and 61/495,540.

Independent claims in plain language

Claim 1 — Grid-tied multiple-channel inverter (apparatus)

A multi-channel microinverter that:

• has at least two DC input channels (one per PV panel) and an AC output port for feeding the grid;
• has a DC-DC boost converter per channel to raise PV voltage to a level suitable for inversion;
• combines all boost outputs in a DC power combiner wired in parallel so the currents sum;
• feeds a DC-AC inverter, whose output passes through a load interface circuit that filters high-frequency switching noise and is injected onto an internal AC powerline where it is paralleled with grid power;
• uses a digital microcontroller to do MPPT (per channel), boost control, DC-AC inversion and grid synchronization, AC current/voltage monitoring, powerline communications, and switching/isolation logic;
• uses a powerline modem (via an interface circuit) for two-way data over the AC line;
• uses a line sensing circuit to detect grid phase and zero-crossing for synchronization;
• uses a solid-state switch to disconnect the internal powerline from the grid when no solar power is being generated;
• derives its own DC power supply from the DC combiner; and
• includes a "message system" connected to the microcontroller that indicates both the inverter's overall status and the status of each individual input channel.

Claim 12 — Off-grid "AC Master" multiple-channel inverter (apparatus)

Same core architecture (multi-channel boost converters → combiner → DC-AC inverter → load interface → internal AC powerline), but adapted for standalone AC loads rather than the grid. The distinguishing elements are:

• a load detector that measures the impedance of the connected AC load;
• a microcontroller that additionally (i) verifies load impedance is within spec, (ii) initially energizes the dead AC powerline, (iii) keeps delivering AC power so other inverters on the same powerline can synchronize to it (the "AC Master" role), (iv) continuously determines whether the load is too large or too small for the system, and (v) shuts down and flags an error if it is;
• a line sensing circuit that checks for pre-existing AC power before startup and monitors for over/under voltage and current;
• a powerline modem for two-way data over the external AC line;
• a power supply fed from the DC combiner; and
• the same message system for overall and per-channel status.

Claim 16 — Method of indicating status

A method comprising: monitoring a multi-channel power inverter and each of its input channels, and actuating LEDs based on that monitoring, wherein a system status LED is controlled to indicate five states: (a) inverter working; (b) working but with warnings; (c) inverter has errors; (d) AC grid has errors; and (e) inverter is off.

Dependent claims at a glance: 2/13 (message system = multiple LEDs + microcontroller-driven LED driver setting color and/or pattern); 3/14 (system status LED five-state set — note 14 recites "AC load has errors" instead of grid); 4/17 (per-channel status LED five-state set: working / low input voltage / errors / warnings / off); 5 (single- or three-phase output); 6/15 (an enclosure containing elements a)–m)); 7 (enclosure hardware: AC input/output grommets, cables, connectors, mounting bracket with slots — see FIG. 1, refs 4–14, 30, 32); 8 (per-channel DC grommet, pair of DC cables, male and female DC connectors — the spec describes these as standard MC4 connectors); 9 (microcontroller main-program loop: on/off based on source/inverter/line conditions, power statistics, diagnostics, per-channel redundancy, unit addressing, powerline comms, answering data-acquisition queries — see FIG. 6); 10 (system-LED subroutine: test a case number and set LED color/pattern accordingly — see FIG. 7/8, Cases 1–5 and 11–15); 11 (per-channel LED subroutine looping over i = 1…m, Cases 6–10 and 16–20).

Litigation posture of the '488 patent

The patent is a high-assertion asset for CyboEnergy. Confirmed and reported matters:

PTAB / IPR: I found no IPR petition directed specifically at the '488 patent. The IPR I did locate — IPR2024-01106 — was filed by Hoymiles USA against CyboEnergy and was terminated by joint motion in October 2024 following a license/settlement agreement (settlement filed as confidential Ex. 1011). Multiple sources tie that dispute to the '133 and '489 patents, but I could not authoritatively confirm which patent claim was challenged; treat any '488-specific IPR assertion as unverified.

CAFC (2026): I found no Federal Circuit docket or appeal for US 9,331,488 in the sources I could reach. My second search on this point hit a tool step-limit before returning results, so I cannot rule out a pending appeal — flagging this as a genuine gap rather than a negative finding.

Cautions and corrections

• Secondary-source errors to beware of: Several Ramey LLP announcements and a Korean IP-NAVI entry mis-describe the '488 patent. The Home Depot press item says '488 is "centered on maximizing power production with low solar radiation" — that is actually US 9,331,489. A Chinese trade article labels '488 "太阳能电池连接装置" (solar cell connection device). The issuing patent text is controlling: '488 is the enclosure + LED message system patent; '133 is the smart/scalable inverter patent; '489 is the low-sunlight power maximization patent.
• Claim 12 limitation: under 35 U.S.C. § 112 ¶4 / AIA § 112(f) practice, "message system" in claims 1/12 may be argued to be functional; the specification supplies structure via LED driver circuit 118/158 plus the FIG. 7/8 subroutines, which is why CyboEnergy has litigated claims 1, 12, and 16 in parallel.
• Assertion style: In the '488 counts, CyboEnergy has generally pled indirect infringement (induced + contributory) alongside direct infringement, on the theory that purchasers' instruction manuals direct daisy-chained installation — a point that drew criticism in the parallel '133 count in the Home Depot complaint analysis.
• A Chinese-language commentary notes the '488 patent has no Chinese counterpart (US-only family), which matches the Google Patents "Country Status (1) US" data.

If you need verification of specific dockets, the primary sources to pull directly are CourtListener docket IDs (e.g., 69444296 for W.D. Tex. 7:24-cv-00319), PACER Monitor case 59719037 for C.D. Cal. 2:25-cv-07964, and the PTAB P-TACTS petition record for IPR2024-01106 — I did not have authenticated access to any of these.

I'll search litigation databases for this specific patent number.

Let me verify additional defendants and check for any cases I may be missing.

Let me check the Generac matter and the Unified Patents caselist for this patent.

I'll verify the structured data's "no PTAB activity" default against live sources, focusing on the IPR2024-01106 lead from the prior section and any other AIA trials involving this patent or its family.

Key correction found: IPR2024-01106 is against US 8,786,133, not the '488 patent. Let me search for any AIA proceeding actually naming 9,331,488, plus family-wide PTAB activity.

Found a second proceeding: IPR2024-00360 (Altenergy v. CyboEnergy). Let me determine which CyboEnergy patent it targets and check for any others.

Proceedings overview

There are zero AIA trial proceedings on US 9,331,488. The USPTO Open Data Portal's structured "PTAB proceedings on file" block for this patent returns no IPR, PGR, or CBM — and my live searches corroborate that: every CyboEnergy AIA petition I could locate in the public record is directed at the family siblings US 8,786,133 ('133) and US 9,331,489 ('489), never at the '488 patent. Claim-count breakdown: 0 canceled / 0 sustained / 0 settled / 0 institution-denied — because nothing was ever filed.

Defensive posture: the patent is unattacked, not hardened. All 17 claims (independent claims 1, 12, 16 plus dependents 2–11, 13–17) remain presumptively valid under 35 U.S.C. § 282 and have never been construed by the Board. That cuts both ways: there is no favorable FWD to quote, but equally there is no § 315(e) estoppel against anyone and no Board claim construction that a defendant would have to live with. The '488 patent is a clean slate for an IPR petitioner, subject only to the § 315(b) one-year time bar described in the Strategic Summary.

⚠️ Correction to the prior section: the earlier analysis could not confirm which patent IPR2024-01106 challenged. It is now confirmed — that petition challenged claims 15–24 of US 8,786,133 B2, not the '488 patent. Nothing in that proceeding touched '488.

⚠️ Do not confuse this patent with same-numbered strangers. Two PTAB records surface "the '488 patent" and are unrelated to CyboEnergy's '488. One is an ex parte/FWD excerpt concerning a swather (agricultural windrower) patent citing Schnittjer, Zhavoronkin, CA Honey and US Honey; the other is an IPR petition excerpt about vehicle charging schedules citing Ferro, Lowenthal and Hafner, with child applications 17/012,325 and 15/848,017. Neither is US 9,331,488 (Enclosure and Message System of Smart and Scalable Power Inverters). A petitioner's obviousness theory imported from those documents would be malpractice.

No proceeding to report — what that means in operative terms

Because there is no petition on file, there is no FWD, no institution decision, no panel, no settlement order, and no appeal for this patent. I will not manufacture one. The nearest-in-family proceedings are set out below expressly as non-'488 context, because they are the best available evidence of (a) what art CyboEnergy's adversaries have already spent money on and (b) how this patent owner behaves when attacked.

IPR2024-01106 — Hoymiles USA, Inc. v. CyboEnergy, Inc. (different patent: US 8,786,133)

• Type: Inter Partes Review
• Patent at issue: US 8,786,133 B2 — not US 9,331,488
• Filed: 2024-06-27 (filing date accorded 2024-07-16)
• Status: Terminated-Settled — "DECISION Settlement Prior to Institution of Trial 37 C.F.R. § 42.74" (2024-10-28). No institution decision was ever made; the panel expressly stated the order "does not constitute a final written decision pursuant to 35 U.S.C. § 318(a)."
• Judge panel: Karl D. Easthom, Monica S. Ullagaddi, and Russell E. Cass, Administrative Patent Judges; opinion authored by Cass, APJ
• Petition grounds: challenged claims 15–24 of the '133 patent (per Patexia's docket summary). Grounds and prior-art specifics are not set out in the termination decision and are not public in the record I retrieved — do not assume grounds.
• Institution decision: None. Termination came pre-institution.
• Final Written Decision: None.
• Settlement / termination: Joint Motion to Terminate filed 2024-10-23 after Board authorization; parties represented that they "have settled their dispute regarding the '133 Patent, including this proceeding and Patent Owner's assertion of the '133 Patent in the related district court litigation against Petitioner." The License and Settlement Agreement (Ex. 1011) was granted confidential treatment under 35 U.S.C. § 317(b) and 37 C.F.R. § 42.74(c) and is designated "Parties and Board Only." Terms are confidential. The Board found good cause to dismiss the Petition and terminate, given the early stage.
• Post-termination housekeeping: Petitioner filed a Request for Refund of Post-Institution Fees (2026-01-20); refund approved 2026-01-27.
• Counsel: Petitioner — Darlene Ghavimi-Alagha and Matthew Blair, K&L Gates LLP. Patent Owner — Gianni Minutoli, Larissa Bifano, Stefan Greenewald, DLA Piper LLP.
• Appeal: None (nothing appealable).
• Defensive value for '488: Essentially zero, and that is the point. Because no FWD issued, no § 315(e)(2) estoppel attached to Hoymiles (or anyone) on the '133 patent, and nothing attached to '488. A defendant cannot borrow this outcome.

Source: IPR2024-01106 Termination Decision (Docket Alarm copy) · Patexia docket summary

IPR2024-00360 — Altenergy Power System USA, Inc. v. Cheng, George Shu-Xing (different patent: US 8,786,133)

• Type: Inter Partes Review
• Patent at issue: US 8,786,133 B2 (Ex. 1001 is the '133 patent; Ex. 1011 is US 9,331,489, cited as context only — no '488 exhibit appears in the list)
• Filed: 2023-12-22 (filing date accorded 2024-01-17)
• Status: Terminated — "Termination Dismissal Prior to Institution of Trial" (2024-03-20), following Petitioner's Unopposed Motion to Withdraw Petition and Refund Request (2024-03-07). Refund approved 2024-04-08.
• Judge panel: Not public — no panel was designated because no institution decision issued.
• Petition grounds: Petitioner filed a Declaration of Mladen Kezunovic (Ex. 1003) and these references: US 2009/0020151 (Fornage), US 2008/0283118 (Rotzoll '118), US 2005/0105224 (Nishi), US 2002/0038667 (Kondo), Kjaer, Review of Single-Phase Grid-Connected Inverters for PV Modules, and US 2009/0000654 (Rotzoll '654). Ex. 1010 is the W.D. Tex. claim-construction decision in CyboEnergy v. Altenergy Power System USA, No. 6:22-cv-01136-KC (Dkt. 44, filed 2023-12-20). The specific statutory basis and the exact set of challenged claims are not stated in the documents I retrieved — I decline to guess.
• Institution decision / FWD: None. Withdrawn and dismissed pre-institution.
• Settlement / termination: Withdrawal was characterized as unopposed; the underlying district-court case, 6:22-cv-01136 (W.D. Tex., filed 2022-10-28), is on the related-case docket. Whether a confidential agreement drove the withdrawal is not stated in the public docket entries I saw.
• Appeal: None.
• Defensive value for '488: None. But note the strategic datum — this is the one petitioner that actually assembled a substantive '133 invalidity record (Fornage/Rotzoll/Nishi/Kondo/Kjaer, plus a Kezunovic declaration). If you are defending against '488 and want a head start on § 103 framing for the inverter architecture, that exhibit set is the most developed public starting point a family adversary has produced. It says nothing about the LED/enclosure claims of '488.

Source: Patexia IPR2024-00360 document list

IPR2024-00362 — Altenergy Power System USA, Inc. v. (Cheng / CyboEnergy) (family context — details unverified)

• A companion Altenergy petition number appears on the Patexia "related cases" list for the Altenergy and Sensata dockets. I could not retrieve its petition, exhibit list, patent, or disposition, and the USPTO ODP block shows nothing for '488. I flag it as an unresolved item rather than assert its subject matter. It does not appear to involve '488 based on the ODP data, but confirm against P-TACTS before relying on that.

Source: Patexia related-cases listing

Strategic summary

Claim status on '488: all 17 claims UNTESTED. No independent claim (1, 12, 16) and no dependent claim (2–11, 13–17) has been canceled, confirmed, or even construed by the Board. Contrast this with the family: the '133 patent drew two petitions (IPR2024-00360, IPR2024-01106), both of which died pre-institution, and the '489 patent was the subject of the W.D. Tex. Markman exercise in 6:22-cv-01136-KC. CyboEnergy has chosen to assert '488 broadly — claims 1–17 in NEP (5:23-cv-06121), Home Depot (7:24-cv-00319), Deye (2:25-cv-07964), and Yotta (1:25-cv-01300); and, more narrowly, only claim 16 against Duracell Power Center (3:24-cv-08891), the sole independent method claim. That Duracell carve-out is worth reading as a signal about which claim CyboEnergy believes is easiest to prove against a brochure-only record.

Estoppel landscape: clean. Section 315(e)(2) estoppel attaches only after a final written decision on a claim. Neither IPR2024-00360 nor IPR2024-01106 produced one, so no petitioner or privy is estopped from raising any ground — including against '488. There is likewise no Board claim construction of "message system," "load interface circuit," "DC power combiner," "system status LED," or "LED driver circuit" to argue about. A defendant today can raise any § 102/§ 103 ground supported by patents or printed publications, subject to § 325(d) discretion if the art was already before the Examiner. Practical caution: the '488 file history and its cited art (58–64 references, including Capstone turbogenerator art, Bower US 2003/0111103, Larankelo US 2008/0283118 / US 2009/0000654, Pfeifer US 2008/0111517, and the Tigo references) will trigger § 325(d) arguments, so a petition built on art already cited starts behind.

Pattern signals. (i) No repeat petitioner on '488 — because there is no petitioner at all. (ii) CyboEnergy litigates aggressively and settles selectively: it sued NEP in 2021, settled 2022-10-19, then sued again in 2023 for breach of the settlement's sell-through term, alleging NEP's counsel said NEP "viewed CyboEnergy's patents as invalid." Hoymiles settled pre-institution in October 2024. Altenergy walked away unopposed in March 2024. Deye dismissed voluntarily (~2025-12-12) and Home Depot dismissed voluntarily (2025-05-19). (iii) No defensive aggregator appears in the chain — no Unified Patents or RPX petition against this family surfaced. (iv) No CAFC appeal exists for '488, and none could: no FWD issued. (v) The core drivers of CyboEnergy's war chest are Ramey LLP (Texas filings), Banie & Ishimoto LLP (Duracell, N.D. Cal.), and DLA Piper as patent-owner PTAB counsel — a materially different quality profile at the PTAB than in the district courts.

Vehicle availability. Because the '488 patent's effective filing date is 2011-06-30 (provisional US 61/503,400) and its application was filed 2012-06-29, it is a pre-AIA patent: PGR is unavailable (PGR requires an effective filing date on or after 2013-03-16). CBM is unavailable — the transitional covered-business-method program sunset on 2020-09-16 and this is not a financial-services patent in any event. IPR is the only AIA vehicle. Ex parte reexamination under 35 U.S.C. § 302 remains available to anyone at any time and is not subject to the § 315(b) bar — worth considering given that no estoppel risk exists yet.

Recommended next steps

1. If you are a defendant facing a '488 demand, the absence of PTAB activity is the headline. State it plainly: there are no proceedings, no FWD, no estoppel, and no Board construction. You are the first mover if you file. The best evidence you can cite is the ODP "no proceedings" result for US 9,331,488 plus the two pre-institution family terminations — IPR2024-01106 termination decision and the IPR2024-00360 docket — showing two adversaries who started family petitions and abandoned them before getting a merits ruling.

2. Run the § 315(b) clock first, before anything else. If you were served with a complaint asserting '488 more than one year ago, you are likely time-barred from filing an IPR on it. Note the Click-to-Call Technologies LP v. Ingenio, Inc., 899 F.3d 1321 (Fed. Cir. 2018) (en banc as to § 315(b)) holding that a complaint later voluntarily dismissed still triggers the bar — which matters for Home Depot (dismissed 2025-05-19) and Deye (dismissed ~2025-12-12) and their customers/privies. Conversely, if the prior case against you was dismissed for lack of jurisdiction, the service does not count. Confirm the service date on your own docket before spending money; I am flagging this as my practice analysis, not as a fact I verified for any particular defendant.

3. If you are inside the § 315(b) window, the trial-stage clock runs fast. Under 35 U.S.C. § 314(b) the Board's institution decision is due within 6 months of a complete petition, and under § 316(a)(11) the FWD is due within 12 months of institution (extendable up to 6 months for good cause). Budget your POPR at ~3 months from the notice of accord and your § 316 deadline pressure accordingly.

4. Attack the claims CyboEnergy actually asserts, not the whole patent. If you face a brochure/manual-only record, note that CyboEnergy asserted only claim 16 against Duracell. Claim 16 is a pure method claim — "monitoring the status of a power inverter having multiple DC input channels and monitoring each of the input channels; and actuating LEDs based on the monitoring" — with the five enumerated system-LED states. Method claims are the classic ground for a § 112(b) indefiniteness and § 101 eligibility attack, and for an on-sale/public-use § 102(a) challenge that requires no printed publication. The specification's own grounding is thin: the five states come from Tables 1 and 3, and the "message system" is functional (LED driver circuit 118/158 plus the FIG. 7/8 subroutines) — the same § 112(f)-versus-structure fault line the prior section flagged.

5. Prioritize the Best Mode / written-description-of-the-enclosure claims (7 and 8) separately. Claims 7–8 recite physical structure — AC input/output grommets, cables, connectors, mounting bracket with slots; and per-channel DC grommet, a pair of DC cables, and a DC male/female connector pair. The specification describes those DC connectors as standard MC4 connectors. That is a strong obviousness posture against a well-known industry-standard part, and the "aluminum case with an alodine corrosion layer" recitation in the spec is likewise conventional.

6. Do not rely on the Ramey LLP press releases or trade-press coverage. The Home Depot announcement mis-describes '488 as "centered on maximizing power production with low solar radiation" — that is US 9,331,489. The issuing patent text is controlling: '488 is the enclosure + LED message system; '133 is the smart/scalable inverter; '489 is the low-sunlight power maximization patent. The IP-NAVI Korean entry, by contrast, gets the '488 identification right (title and IPC H02J 1/10, H02M 7/48, H02J 3/38, H02M 7/44, H02M 1/00), though it renders claim scope in Chinese-character form — use it for bibliographic corroboration only.

7. Verify the unresolved item before you file anything. Pull P-TACTS directly for IPR2024-00362 to confirm its target patent and disposition, and pull the full IPR2024-01106 and IPR2024-00360 petition papers if you want the grounds verbatim. In both instances the Board's own documents say the proceedings ended without reaching the merits, so neither is citable as an invalidity finding — only as evidence that two competitors tried and quit.

I'll research the assignment record for this patent. Let me run several targeted searches.

Let me dig into the assignment record specifics and any post-issuance transfers.

Let me verify whether any post-issuance transfers exist and pin down the correspondent of record and CyboEnergy's corporate status.

Assignment record reconstruction — US 9,331,488 B2

Source access limitation (stated up front): I do not have authenticated access to USPTO Assignment Center (assignmentcenter.uspto.gov) or to the legacy Assignment Search at assignment.uspto.gov/patent/index.html. The assignment data below is taken from the Google Patents "Legal Events" record for US 9,331,488 B2 (which mirrors the USPTO assignment record and reproduces the reel/frame), corroborated against Justia Patents assignee listings and CyboEnergy corporate/product documents. I could not retrieve the recorded correspondent field for the assignment itself — see §3 below for what I could and could not establish. Today's date in my system is 2026-09-19; consistent with the prior section's flagged date discrepancy.

Inventors

Unusual patterns — checked, not found: The classic red flag ("all inventors leave the original assignee within 12 months of filing, preceding a fire-sale") is not present. Two of three inventors (Mulkey, Cheng) remained at CyboEnergy and kept inventing for it for another decade — e.g., US 11,258,267 (filed 2019-08-22) and US 12,424,854 (filed 2022-01-13) both list Cheng and Mulkey. Martinez's single-appearance profile is notable but is a scope-of-contribution observation, not a departure signal.

One timing quirk worth recording: the assignment's stated effective/execution date is 2011-07-15, i.e. ~15 days after the provisional (US 61/503,400, filed 2011-06-30) but ~11.5 months before the non-provisional was filed (2012-06-29). This is an ordinary startup sequence (provisional filed first, formal inventor assignment papered shortly after).

Original assignee

CyboEnergy, Inc. (recorded as "CYBOENERGY, INC., CALIFORNIA"; corporate address Rancho Cordova, CA).

• Ships products embodying the claims — yes. CyboEnergy sells the CyboInverter line under part numbers including CIM-1000N, CIM-1200H, CIM-1200Na / CIM-1200Ya, and the daisy-chained CIM-5000Z quad pack (Master Z1 + three Slaves). Product literature describes the multi-channel architecture recited in the claims directly: "Each CyboInverter has 4 input channels with MPPT for each solar panel," "Master/Slave daisy-chain," "AC Input from the Grid or AC generator," 120 V/240 V single-phase. Marketing materials state units are "patented, UL1741 certified, NEMA6 (IP67) rated and made in the USA." (CyboEnergy Residential Solar System Design Guide, Rev 1.1, May 2024; CIM-5000Z spec sheet Rev 6.6, Jan 2024.)
• Primary line of business: design/manufacture of smart, scalable on-grid and off-grid solar power inverters; CyboEnergy also explicitly offers IP licensing, assembly and private-label arrangements to strategic partners (CyboEnergy press release, Apr 12 2022).
• Corporate structure: stated to be "an affiliate of CyboSoft, General Cybernation Group Inc." — a privately held California group, therefore no SEC 10-K/8-K cross-check is possible (not a public filer). I found no bankruptcy, dissolution, or Chapter 7/11 record.
• Current status: operating. It is still the listed current assignee on Google Patents; it is paying maintenance fees (4th-year fee, small-entity status); it obtained a new patent as recently as 2025-09-23 (US 12,424,854); and it is actively litigating in its own name as plaintiff (Northern Electric Power 2021; Home Depot and Duracell 2024; Deye and Yotta 2025; Generac 2026 per the prior section).

Assignment timeline

Exactly one assignment is recorded against US 9,331,488. There is no post-issuance transfer of any kind — no licence recordation, no security agreement, no change of name, no merger, no release, no correction.

2011-07-15 (executed; stated "Effective date") / recorded 2012-07-20 — Reel 028594/0003

• Conveyance: Assignment — "ASSIGNMENT OF ASSIGNORS INTEREST."
• Assignor: MULKEY, STEVEN L.; CHENG, GEORGE SHU-XING; MARTINEZ, MANUEL (all three joint inventors)
• Assignee: CYBOENERGY, INC. (California)
• Correspondent: Not verifiable from my sources. I could not retrieve the recording correspondent field. The attorney/agent of record printed on CyboEnergy's later patents in the same family is Donald E. Stout, Stout, Uxa & Buyan, LLP (see, e.g., US 9,871,379 B2 face: "(74) Attorney, Agent, or Firm — Donald E. Stout; Stout, Uxa & Buyan, LLP"). This is the firm most likely to have filed the recording, but I am not asserting it as the recorded correspondent without seeing the Assignment Center field. Flagging rather than inferring, per the no-fabrication constraint.
• Context: initial inventor-to-company assignment — the standard equity/obligation paper for a startup, executed around the provisional filing. Not a fire-sale, not a reorg, not a transfer to an asserter.

Other recorded legal events (Google Patents legal-events mirror), none of which are assignments:

• 2012-07-20 — AS / Assignment (the link above)
• 2016-05-03 — STCF / patent grant ("PATENTED CASE")
• MAFP — "PAYMENT OF MAINTENANCE FEE, 4TH YR, SMALL ENTITY" (year of fee payment not itemized in the extract I retrieved)

Practical conclusion: the chain is a single link, and it terminates at the original operating-company assignee. For a patent with this much assertion activity (six-plus suits), the absence of any recorded ownership change is itself the finding — CyboEnergy is litigating as the original owner, not as a transferee.

Timeline diagram

timeline
    title Ownership of US 9331488
    2011 : Provisional app filed Jun 30
         : Inventors assign rights to CyboEnergy Jul 15
    2012 : Non-provisional filed Jun 29
         : Assignment recorded Jul 20 Reel 028594
    2013 : Pre-grant publication Jan 3
    2016 : Patent issued May 3
    2021 : First infringement suit filed
    2024 : Suits filed vs Home Depot and Duracell
    2025 : Suits filed vs Deye and Yotta

NPE / troll-pattern signals

Neutral observations that cut against a clean "pure operating company" story (recorded for balance, not as NPE signals): CyboEnergy is small (small-entity fee status), explicitly markets an "IP licensing, assembly, and private label" business line, and its docket shows at least six filings in five years, several of which were voluntarily dismissed (Home Depot May 2025; Deye ~Dec 2025). That is a hybrid operating-company-plus-assertion profile — but hybrid assertion by a company that actually ships the patented product is not the NPE pattern, and no assignment evidence converts it into one.

Verdict

Operating-company assertion.

The entire recorded chain is a single inventor-to-company assignment — Reel 028594/0003, executed 2011-07-15, recorded 2012-07-20 — conveying the '488 patent from Mulkey, Cheng and Martinez to CyboEnergy, Inc., and nothing has been recorded since. CyboEnergy still owns it, still pays maintenance fees on it (small entity), still ships CyboInverters that practice the multiple-input-channel / daisy-chain / AC-input architecture recited in claims 1 and 12, and sues named competitors and resellers (Northern Electric Power, Duracell Power Center, Home Depot, Deye, Yotta, Generac). With no shell-entity transfer, no known asserter, no cascading transfers, no pre-suit transfer, no bankruptcy sale, no privateering, and no defensive-aggregator terminus, none of the eight NPE signals is present; the only mild anomalies — a boutique prosecution firm recurring across the family and a parallel "IP licensing" revenue line — are not evidence of an NPE chain.

Verify directly at: USPTO Assignment Center (search patent number 9331488 → confirm reel 028594/0003 and read the correspondent field, which I could not retrieve) and the mirror at https://assignment.uspto.gov/patent/index.html. Reel/frame corroboration is also visible in the Google Patents legal events for US 9,331,488 B2 (AS | Assignment | Owner name: CYBOENERGY, INC., CALIFORNIA ... REEL/FRAME:028594/0003).

Cross-reference note vs. the prior section: the prior section's figure of "assignment recorded 2012-07-20, Reel/Frame 028594/0003" is consistent with what I independently recovered here; I found no contradiction. The one thing I could add is the effective date of 2011-07-15, which the prior section did not state — treatment of the assignment as pre-filing paper is correct, but it post-dates the 2011-06-30 provisional by two weeks rather than preceding it.

I'll ground this in live sources — first confirming the patent's own citation list and then pulling bibliographic detail on the most probative cited references.

Let me pull bibliographic detail on the remaining probative cited references, including the ones most relevant to the LED/status-display and multi-input-combiner aspects.

Prior-Art Analysis — US 9,331,488 B2

Scope discipline first. Confirmed against the authoritative full text: the patent is US 9,331,488 B2, "Enclosure and Message System of Smart and Scalable Power Inverters," Application No. 13/537,206, filed 2012-06-29, granted 2016-05-03, priority 2011-06-30 (Prov. 61/503,400), assignee CyboEnergy, Inc. Everything below keys to that literal ID. I have excluded the neighboring CyboEnergy patents that share the family and the litigation record — US 9,331,489 (low-sunlight power maximization), US 9,253,916, US 9,093,902, US 8,786,133, US 9,176,270 — and the related US 2013/0002031 A1 publication. Note that the Google Patents text lists US 2013/0002031 A1 among the '488's own cited documents, which is an artifact of how the citation tables are assembled; it is the '488's own pre-grant publication and cannot be prior art against itself.

Sourcing caveat (honest). I do not have authenticated USPTO PatentCenter/PACER/P-TACTS access in this session. My identity confirmation and the citation lists come from the Google Patents full text supplied in your message plus corroborating mirrors (patentimages PDF of the granted patent; Justia's patents.justia.com/patent/9331488 citation table; patentsencyclopedia, SumoBrain, RPX Insight, Unified Patents portal). Where I characterize a reference's disclosure from its title or abstract rather than its full text, I say so and mark confidence. Where I could not fetch a reference at all, I say that too.

1. The § 102 framework that governs this analysis

The '488 was filed 2012-06-29, i.e., before the AIA first-inventor-to-file provisions took effect on 2013-03-16. Pre-AIA 35 U.S.C. § 102 therefore controls (post-AIA § 102(a)(1)/(a)(2) does not apply to this application). Consequences that matter a great deal for this exercise:

The single most important takeaway: the examiner's citation list on the face of the '488 mixes (i) genuine pre-2011 art, (ii) the applicant's own related applications, and (iii) a large block of references that published after 2011-06-30 and are therefore not § 102(a)/(b) art at all. Treating the whole list as "prior art" would be a category error; only a subset can legally anticipate anything.

2. Triage of the citation list on the face of the '488

The patent's front page carries two citation tables whose counts disagree — "Citations (58)" and "Patent Citations (64)" — which I flag as a document inconsistency (the extra entries in the 64-list are additional Capstone Turbine and Cheng control patents). The forward-citation table ("Families Citing this family (34)") is not prior art and is ignored here.

2a. The examiner's own references ("cited by examiner," asterisked on the face)

Google's convention in these tables is * = cited by examiner, † = cited by third party. Only five documents in the list carry the examiner asterisk:

This five-document set is the clearest signal of what the examiner viewed as the closest art to the message-system limitation — and notably, none of them is a multi-channel PV inverter with per-input-channel LED status.

2b. Pre-2011-06-30 references that are substantively probative

US 2011/0012430 A1 (Cheng & Mulkey; General Cybernation Group, Inc.), Smart and scalable power inverters — pub. 2011-01-20, filed 2010-07-15, App. 12/837,162, granted as US 8,786,133 B2. This is the application expressly incorporated by reference into the '488 specification. US 2003/0111103 A1 (Bower, Thomas, Ruby; Sandia) — pub. 2003-06-19, priority 2001-10-25, WO 03/036688. US 2008/0283118 A1 (Rotzoll & Kapur; Larankelo) — pub. 2008-11-20, PCT filed 2008-05-16, prov. 60/938,663 (2007-05-17). US 2009/0000654 A1 (Rotzoll, Kapur, Patil; Larankelo) — pub. 2009-01-01, filed 2008-05-15 (12/121,616). Plus the multi-source / MPPT / boost-converter art: US RE37,126 E1 (Peng; 2001-04-03), JP H09-275637 A (Sanyo; 1997-10-21), JP 2001-008383 A (Nissin Electric; 2001-01-12), US 2004/0264225 A1 (Bhavaraju/Ballard; 2004-12-30), US 2010/0202177 A1 (Kajouke/GM; 2010-08-12), US 2008/0111517 A1 (Pfeifer; 2008-05-15), US 2009/0159113 A1 (Morimoto), US 2009/0160258 A1 (Allen), US 2009/0174259 A1 (Lin), US 2010/0231045 A1 (Collins/First Solar), US 2010/0237703 A1 (Stern), US 2011/0148195 A1 (Lee/Samsung SDI; pub. 2011-06-23, seven days pre-priority).

2c. References cited that are later than the priority date and thus not § 102(a)/(b) art

US 2012/0212065 A1 (Cheng; 2012-08-23) and US 2012/0313443 A1 (CyboEnergy; 2012-12-13) — these are the 61/442,991 (redundancy) and 61/495,540 (off-grid AC Master) families also incorporated by reference; US 2014/0252862 A1 and US 2014/0265584 A1 (CyboEnergy) — the applicant's own later filings; US 2012/0025618 A1 (Erickson/Miasole, 2012-02-02), US 2012/0104872 A1 (Marroquin, 2012-05-03), US 2012/0138123 A1 (Newdoll, 2012-06-07), US 2011/0273022 A1 (Dennis, 2011-11-10), and the three Tigo grants US 8,102,074 B2 (2012-01-24), US 8,860,241 B2 (2014-10-14), US 8,933,321 B2 (2015-01-13). ⚠ Nuance: a document that published after 2011-06-30 can still be § 102(e) art if its US filing date precedes the applicant's invention date — so the Tigo grants, the Miasole/Marroquin/Newdoll/Dennis publications, and E4 above are § 102(e) candidates, not § 102(a)/(b) art. I flag this because conflating the two is the commonest error in '488 prior-art write-ups (the same error that produced the Ramey LLP / IP-NAVI mis-attributions already noted in the earlier litigation sections).

2d. Background clusters with essentially no anticipation value

The Capstone Turbine cluster — US 5,903,116; 6,031,294; 6,192,668; 6,265,786; 6,325,142; 6,487,096; 6,495,929; 6,612,112; 6,784,565; 6,870,279; 6,958,550; 6,960,840; plus pubs 2001/0052704, 2002/0099476, 2002/0166324, 2003/0007369, 2004/0119291, 2004/0135436 (all 1997–2005) — discloses turbogenerator/motor controller and power controller technology. None discloses per-channel PV DC-DC boost converters, a DC power combiner, a powerline modem, or an LED message system. The Cheng control patents — US 6,055,524; 6,360,131; 6,556,980; 6,684,112; 6,684,115; 7,016,743; 7,142,626; 7,152,052; 7,415,446 (1997–2008) — are the inventor's own Model-Free Adaptive (MFA) control work, relevant at most to claim element 1(h). Both clusters read as IDS background, not as the art that forced claim narrowing.

3. Reference-by-reference: citation, dates, description, and § 102 mapping

Methodological rule I apply below: anticipation under § 102 requires all elements of a claim in a single reference, arranged as claimed (Verdegaal; Net MoneyIN). I therefore state for each reference (a) which elements it discloses, and (b) whether it could anticipate a whole claim — and I do not label anything anticipatory unless the reference discloses the entire claimed combination.

3.1 Ranked shortlist — the most relevant cited prior art

3.2 The incorporated-by-reference family — a § 102 problem in itself

The '488 specification states verbatim: "In the U.S. patent application Ser. No. 12/837,162 … we described the novel Smart and Scalable Power Inverters…," and also incorporates Prov. 61/442,991 (redundancy) and Prov. 61/495,540 (off-grid AC Master). Two consequences:

• US 2011/0012430 A1 / US 8,786,133 cannot be treated as § 102 art against the subject matter it contributes, because that subject matter is part of the '488's own disclosure by incorporation (MPEP 2132.01 logic). It also raises the "by another" question under pre-AIA § 102(a)/(e): the parent's inventive entity is Cheng + Mulkey, while the '488's is Mulkey + Cheng + Martinez — a different entity, which is exactly the fact pattern that generates litigation-grade inventorship argument. Practically, the parent is best used as a § 103 (obviousness) reference and a § 112 written-description anchor, not as an anticipation reference.
• The redundancy limitation in claim 9(d) ("run redundancy routine for every input channel") traces to 61/442,991, and the entire off-grid claim 12 architecture traces to 61/495,540 — neither of which is § 102 art. That is a point in the patent's favor on the two independent apparatus claims.

3.3 Non-patent literature cited (all pre-2011, so § 102(b)-eligible in principle)

4. Bottom line on § 102

No reference cited on the face of US 9,331,488 anticipates any of independent claims 1, 12, or 16.

• Claim 1 requires, in a single multi-channel unit, the DC-DC-boost-per-channel → parallel DC power combiner → DC-AC inverter → high-frequency-filtering load interface → internal AC powerline → digital microcontroller with per-channel MPPT and powerline communications → powerline modem → phase/zero-crossing line sensing → solid-state switch → power supply fed from the DC combiner → message system indicating both overall status and the status of each input channel. The closest reference, US 2011/0012430 A1, supplies essentially everything through element (k) but is missing the message system and describes the supply as tapping the external AC powerline rather than the DC combiner (element (l)). The examiner-cited art (US 2007/0040532 A1, US 2011/0231456 A1) reaches status/error indication only in non-analogous, non-multi-channel contexts, and E4 post-dates the priority date. So claim 1's novelty rests on (l) and (m) — and (m) is the patent's actual point of novelty, which matches the specification's own summary and the reason the patent issued.
• Claim 12 adds the load-impedance detector, the dead-line energize step, the AC-Master synchronization-of-other-inverters role, and the too-large/too-small load shutdown. No cited reference discloses that combination. JP H09-275637 touches parallel source operation, and US 2003/0111103 A1 mentions stand-alone/hybrid systems, but neither reaches the AC-Master role or the impedance test.
• Claim 16 (method: monitor the multi-channel inverter and each input channel, then actuate LEDs across five enumerated system states) is the farthest from all cited art. Nothing cited discloses per-input-channel optical status indication with a defined color/pattern state vocabulary. The dependent claims built on it (2, 3, 4, 10, 11, 13, 14, 17) share that gap; this is consistent with the earlier section's observation that CyboEnergy litigates claim 16 as the lead assertion against Duracell.
• Dependent claims most at risk are the enclosure/mounting claims 6–8, where the Larankelo references (US 2008/0283118 A1, US 2009/0000654 A1) and Bower US 2003/0111103 A1 supply a mounting bracket on the back of a PV module, a metal/unitary housing, AC cables and connectors, and a parallel daisy-chained AC bus. Even there, claim 8's per-channel pair of DC cables terminating in a DC male and DC female connector (specifically the standard MC4 pair) is not clearly shown by any cited reference and is the strongest part of claim 8.

In short: the citations on the '488 are heavily populated by the applicant's own family and by general-purpose converter/control background, with the genuinely consequential third-party art being the Larankelo mount/gateway pair, Bower's AC PV building block, US RE37,126, and the Sanyo parallel-operation reference — all of which attack the inverter/enclosure architecture that the '488 inherited from its parent, and none of which touches the LED message system that the '488 actually claims.

5. Caveats, gaps, and flagged inconsistencies

1. Full texts not verified this session for: US RE37,126 E1, JP H09-275637 A, JP 2001-008383 A, US 2010/0202177 A1, US 2004/0264225 A1, US 2001/0043052 A1 / US 6,577,098 B2, US 7,710,077 B2, and US 2011/0231456 A1. My element mappings for those are drawn from titles, abstracts and family data. Before relying on any of them, pull the front page and claim 1 of each from PatentCenter.
2. US 2008/0285317 A1 (Rotzoll; Photovoltaic Module-Mounted AC Inverter; Larankelo, filed 2008-05-15) is a companion of the cited US 2008/0283118 A1 and, on its face, discloses a boost converter + buck inverter + PWM control + per-module MPPT + powerline communications + grid-limit shutdown. It is the single most on-point third-party disclosure I saw for elements 1(c), (e), (g), (h), (i), (j) — but it does not appear in the '488's citation list, so I have not counted it as cited art. If you are building an invalidity position rather than an IDS audit, this is the first non-cited reference to pull.
3. "By another" / common ownership is the pivotal unresolved question for US 2011/0012430 A1 (and for US 2012/0212065 A1 and US 2012/0313443 A1): differing inventive entities versus § 103(c)(1) common ownership (original assignee CyboEnergy, Inc.; the parent named General Cybernation Group, Inc.). I cannot resolve inventorship/assignment priority from the data I have.
4. Document inconsistencies to note in the record: (i) the two citation tables on the '488 are titled "Citations (58)" and "Patent Citations (64)" and do not agree; (ii) the '488's own publication US 2013/0002031 A1 is listed among its cited documents; (iii) "General Cyberation Group, Inc." appears in one entry beside the correct "General Cybernation Group, Inc."; (iv) the fetched "Definitions" text renders the Power Electronics Handbook editor as "Arabic H. Rashid" whereas the specification body and the NPL list say "Muhammad H. Rashid" (transcription artifact); (v) the fetched text renders "male AC connector" as "mare AC connector" and "MPPT" as "MDPT" in the off-grid passage (OCR/transcription artifacts, not substantive); (vi) the assignee address reads Rancho Cordova, CA in the granted PDF, not Sacramento as stated in the earlier bibliographic section — a minor correction to that table.
5. Date discrepancy carried forward. The harness prompt states the current date as April 26, 2026 while my system clock reads September 19, 2026, and some litigation datapoints returned by search carry post-April-2026 timestamps. This § 102 analysis is date-independent (it turns on 2011–2016 dates), so the discrepancy does not affect it — but it remains unresolved for the litigation sections and should be reconciled before that work product is finalized.
6. Genuine gap: I have no authenticated access to PatentCenter's "References Cited" / examiner's Form 892 for this application, so my identification of E1–E5 as the examiner's actual five cited references rests on Google Patents' asterisk convention, not on the official 892. Confirming that list from the file wrapper is the highest-value next action for a rigorous prior-art audit.

I'll ground this obviousness analysis in the actual cited prior-art references before building the combinations.

Now let me ground the secondary references — the ones that supply the "message system"/status-indicator and off-grid elements.

Obviousness Analysis — US 9,331,488 B2 under 35 U.S.C. § 103

Building on the prior sections (bibliography, plain-language claim charts, litigation posture). I do not repeat those; I extend them with the §103 work. One flag carried forward: the earlier section noted a date mismatch (prompt date vs. system/search timestamps); the "Prior Art" record I use below is the citation set printed on the face of the '488 and is date-independent.

1. Governing framework and the real battleground

The '488 was filed 2012-06-29 with priority to provisional 61/503,400 (2011-06-30). It is therefore a pre-AIA patent: pre-AIA §§ 102/103 apply, including the pre-AIA § 103(c) common-ownership disqualification discussed in §7 below.

Standard Graham v. John Deere analysis. The single most important structural fact for this analysis is the one identified in the earlier claim section, now with a sharper edge:

Every element of claim 1 (a)–(l) and of claim 12 (a)–(l) is disclosed in US 2011/0012430 A1 — and the '488 specification says so.

The '488 specification states: "In the U.S. patent application Ser. No. 12/837,162, the entirety of which is hereby incorporated by reference, we described the novel Smart and Scalable Power Inverters…" (https://patents.justia.com/patent/[9331488](/patent/9331488)). US 2011/0012430 A1 is the publication of Ser. No. 12/837,162 (General Cybernation Group, published 2011-01-20). Its claim 17 reads, in relevant part:

"…a) at least two DC power input ports… c) for each DC power source, a DC-DC boost converter… d) a DC power combiner connected to said DC-DC boost converters for combining the DC output from all DC-DC boost converters and allowing the said DC-DC boost converters to connect in parallel so that all DC currents are added together; e) a DC-AC inverter connected to said DC power combiner… f) an internal AC powerline that combines the generated AC power with the external AC power from the power grid; g) a load interface circuit… arranged to filter high-frequency components out of the said DC-AC inverter's AC output; h) an MFA microcontroller… monitor the DC boost voltage, control the DC-DC boost converter, perform maximum power point tracking (MPPT), perform DC-AC inversion and AC power synchronization, monitor AC current and voltage… perform powerline communications, and perform logic controls such as AC powerline switching and isolation; i) a powerline modem connected to said microcontroller and said internal AC powerline through an interface circuitry…; j) a line sensing circuit… for detecting the phase and zero-crossing point of the incoming AC power from the power grid; and k) a solid state switch… arranged to disconnect said internal AC powerline from the AC grid during the non-generation time."

That is element-for-element the body of '488 claim 1, in nearly identical wording. The same document's claim 24 recites the main-program loop as "(a) turn on and off the inverter's generation circuit…; (b) calculate the inverter's power statistics…; (c) perform diagnostics…; (d) set the inverter's unit address; (e) perform powerline communications; and (f) respond to queries from data gathering or acquisition devices to report the power statistics" — which is '488 claim 9, less the redundancy step.

Consequence: the §103 dispute is not about the inverter architecture. It is about (i) the "message system" element (claim 1(m), 12(m)) and its LED implementation (claims 2–4, 13–14, 16–17), (ii) the enclosure (claims 6–8, 15), and (iii) the off-grid/AC-Master behaviors (claim 12). I analyze each.

Gap notes I could not close with verified text: (1) '488 claim 1(l) requires the internal DC power supply to be fed from the DC combiner, whereas '430 ¶¶[0044]/[0049] describe the supply as tapped from the external AC powerline ("The external AC powerline 172/232 is tapped by a power supply"). This is a genuine structural difference requiring a secondary teaching or a routine-design argument (see §4.3). (2) I did not verify '430's disclosure of a per-channel status indicator; as far as the record I could retrieve shows, it has none — which is precisely why the combination below is needed.

2. Level of ordinary skill and analogous art

PHOSITA: a B.S. in electrical engineering (or equivalent) with 2–4 years of experience in power-electronics conversion, including grid-tied PV inverters, MPPT, PWM inversion, and microcontroller-based control; capable of routine firmware design and enclosure/mechanical integration for outdoor electronics.

Field of the art: power conversion and, more specifically, distributed/PV-module-mounted inverter systems, their enclosures, and their status-indication/diagnostics interfaces. The "message system" art is not confined to PV: the Federal Circuit and the Board routinely treat indicator/diagnostic subsystems that report the operational state of a power converter as within the same field of endeavor. Precedent supports reading a broad field where the problem to be solved (telling a technician what is wrong with a converter installation) is the same.

3. The prior-art references I verified

References I did not verify in this pass and therefore do not rely on for specific disclosures (they appear on the face of the '488 per the Justia/Google citation tables, with the assignees/inventors shown there): JP H09-275637 (Sanyo Electric, "Fuel cell parallel operation system"); JP 2001-008383 (Nissin Electric, "Photovoltaic power generating set"); US RE37,126 E1 (Peng/Lockheed Martin, "Multilevel cascade voltage source inverter with separate DC sources"); US 2009/0160258 A1 (Allen et al., "Advanced Renewable Energy Harvesting"); US 2009/0159113 A1 (Morimoto); US 2009/0174259 A1 (Lin); US 2010/0202177 A1 (Kajouke/GM, "Voltage link control of a DC-AC boost converter system"); US 2004/0264225 A1 (Bhavaraju/Ballard, MPPT determination); US 2008/0111517 A1 (Pfeifer, charge controller for DC-DC conversion); US 2012/0104872 A1 (Marroquin); US 2009/0000654 A1 (Rotzoll, "Distributed inverter and intelligent gateway"); US 2011/0273022 A1 (Dennis). These are identified by title/assignee only; treat each proposed use of them as needing verification against the specification text.

4. Proposed ground-by-ground combinations

4.1 Ground A — Claims 1–5 (grid-tied multi-channel inverter with message system)

Combination: US 2011/0012430 A1 (primary) + US 2007/0040532 A1 (Bae/Samsung) + US 7,710,077 B2 (HDM), optionally + US 2008/0283118 A1 (Rotzoll/Larankelo).

Element-by-element: (a)–(l) from '430 (§1 above). Elements (b) "AC output port to the grid," (g) "filter high-frequency components," (j) "phase and zero-crossing," (k) "solid-state switch disconnect during non-generation" are literal recitations of '430 claim 17. (m) — the message system — comes from Bae and/or HDM: Bae discloses a microcomputer that determines an inverter error and drives a display part (expressly "an LED… to display information through a light") to display error and error details and a warning message; HDM discloses an inverter/charger whose LED changes color and blink cadence (red / slow-blinking orange-red / fast-blinking orange-red / blank) to announce operating state.

Motivation (this is the crux, and it is unusually strong):

1. Express problem statement in the prior art. Rotzoll ('118) frames the exact problem the '488 solves: with module-mounted distributed inverters, "the identification of that singular PV module failure… is extremely difficult as there is no outright indication that a PV module has failed. Generally the only indication of a PV module failure is a decrease in power production." KSR requires an articulated reason, and a reference's own identification of the unsolved problem is about as direct as it gets. Rotzoll then notes the service consequence — "To do such a repair… the entire PV system must be taken off line." A PHOSITA reading Rotzoll is expressly taught that distributed-inverter installations need an outright indication at the module.
2. The multi-channel architecture multiplies the problem. '430's own cost rationale is that "each smart multi-input power inverter can handle m DC power sources… much more cost effective than the one-on-one design." But a multi-input unit collapses m sources behind one enclosure face; the installer can no longer infer which input is bad from physical location. One indicator per channel is the natural, near-inevitable response.
3. The '488 specification itself supplies the market-pressure rationale (§"To summarize"): the Mini-Inverters "have to work in a harsh environment for a prolonged period of time," and installation cost "accounts for a big percentage of the total cost for a solar power system," so an LED message system "help[s] the installers troubleshoot." That is an admitted design-driver, and it is a driver that existed at the priority date.
4. Color + pattern as the reporting channel was already established. HDM's solid vs. slow-blink vs. fast-blink LED vocabulary, applied to an inverter's operating states, is the same vocabulary '488 Tables 1–4 use (solid green / flashing green / solid red / flashing red / off). KSR § IV: "if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious."
5. Predictable variation / design choice. Moving from Bae's generic "display part (LCD or LED)" to a 2-color LED, and from one system LED to m channel LEDs, is a predictable UI choice with a known benefit (channel-level fault isolation) and no change in principle of operation.

Claim 2 (plurality of LEDs + LED driver circuit controlled by the microcontroller setting color and/or pattern): Bae's microcomputer-driven display part + HDM's LED color/cadence states. A current-driver stage is routine (the '488 spec concedes "any of a number of well known current drivers such as an emitter follower transistor driver").

Claim 3 / claim 16 (five-state system-status set: working / working with warnings / errors / grid errors / off): maps onto Table 1 Cases 1–5. Bae supplies working-vs-error-vs-warning; the grid-fault and off/waking-up states are the natural content of the '430 line-sensing and solid-state-switch logic (which already detects grid phase/zero-crossing and already isolates during non-generation — i.e., it already knows grid status and generation status). Converting known internal states into known external indications is the definition of a predictable combination.

Claim 4 / claim 17 (per-channel set: working / low input voltage / errors / warnings / off): '430 already monitors per-source boost voltage and performs per-source MPPT (claim 17(c), (h)) — it necessarily has per-channel voltage data. Displaying per-channel voltage states via the Bae/HDM indicator paradigm is a direct application.

Claim 5 (single- or three-phase): expressly in '430 claims 3/6/7/10/11/15/19 (single-phase and three-phase, "voltage larger than 240V AC" / "208V AC").

4.2 Ground B — Claims 6–8, 15 (the enclosure)

Combination: US 2008/0283118 A1 (Rotzoll/Larankelo) + US 7,710,077 B2 (HDM) + US 2011/0012430 A1.

• Claim 6/15 ("elements a) through m) disposed in or on an enclosure"): trivially met by any packaged inverter; Rotzoll and HDM each disclose an inverter enclosure.
• Claim 7 (metal case + system status LED + AC input grommet/cable/connector + AC output grommet/cable/connector + mounting bracket with mounting slots):
  • Rotzoll: inverter mounting bracket attached to the back of the PV module, inverter coupled to the bracket, AC cables/connectors for daisy-chain interconnection of adjacent module-mounted inverters, metallic enclosure.
  • HDM: extruded-aluminum case with "three glands… for the AC power input, AC power output, and remote control cables" (i.e., grommet + cable passthroughs, exactly the claim 7(a)–(f) structure), plus through-bolt mounting to a support base/frame (claim 7(g) mounting structure). The '488 spec itself describes alodine surface treatment for corrosion — routine outdoor-enclosure practice.
• Claim 8 (per channel: DC grommet, pair of DC cables, DC female connector, DC male connector): Rotzoll discloses per-module DC input cables with mating connectors; the '488 specification admits the connector choice is dictated by industry standard — "most solar panels are shipped with a pair of standard MC4 DC connectors with DC cables. Therefore, the Mini-Inverter enclosure is designed to include a pair of standard MC4 DC connectors for each input channel." Selecting MC4 male/female pairs (and, per the earlier claim section, the '488 also recites a pair of AC connectors with male output / female input) is an admitted off-the-shelf, standards-driven choice. Per-channel grommets/cable-pairs are the mechanical restatement of "one input channel per panel."

Motivation: Rotzoll expressly seeks an enclosure that fits within the PV module frame depth, minimizes wiring and is field-replaceable; HDM expressly seeks a watertight outdoor package ("mobile applications, such as marine and vehicle, require watertight power devices for use outdoors"). Both are the same problems the '488 states ("harsh environment," installation-cost reduction). Combining a module-back mounting bracket and daisy-chain AC connectors (Rotzoll) with a sealed multi-gland metal case (HDM) is a predictable assembly of known enclosure elements; the arrangement of grommets, cables and connectors about the case is routine mechanical design.

Caveat: claims 6–8 are the weakest target for §103 in a vacuum, because "obviousness" of an enclosure layout is heavily fact-dependent and the patentee will argue that no reference shows the specific spatial arrangement, and that the examiner allowed over Rotzoll (which was of record).

4.3 Ground C — Claim 9 (microcontroller main program)

Combination: US 2011/0012430 A1 + US 2008/0285317 A1 (Larankelo) + US 2011/0231456 A1 (Choi/LS Industrial).

Claim 9 steps (a) turn generation on/off based on DC source/inverter/AC powerline conditions; (b) power statistics; (c) diagnostics; (d) redundancy routine for every input channel; (e) set unit address; (f) powerline communications; (g) respond to queries from data-gathering devices — are almost wholly disclosed by '430 claim 24 (steps a, b, c, e, f, g in the same terms). The only unmatched step is (d), the per-channel redundancy routine, which is described at length in the applicant's own US 2012/0212065 A1 (Cheng, "Scalable and redundant mini-inverters") — see the §103(c) caveat in §7, and note that redundancy (a bypass/reconfiguration of a failed channel) is a conventional reliability technique in modular power conversion (parallel-redundant converter modules). Rotzoll '317 supplies the "local identification code" / address-based query-and-report behavior and self-test; Choi supplies per-inverter inherent ID numbers used as addressing keys for reading and writing parameters of a plurality of inverters.

Motivation: paralleled/daisy-chained modular converters present the known problem of addressing individual units and reporting per-unit data; Rotzoll and Choi both address it; adding a per-channel fault-isolation/bypass step to an inverter that already performs per-channel MPPT and per-channel monitoring (as '430 claim 17(h) does) is a predictable program-flow extension.

One open gap to prove: the DC-combiner-fed internal power supply of claim 1(l) (see §1 gap note). Candidate teachings: US 2010/0202177 A1 (Kajouke/GM, DC-link voltage control) or US 2012/0025618 A1 (Erickson/Miasole, thin-film PV element with integrated DC-DC converter) — both unverified in this pass. If no express teaching is found, the fallback is routine-design: deriving a logic bias supply from the DC bus is a ubiquitous practice, and the '488 spec admits "[o]ther modules… including… DC power supply can be implemented using one or more known combinations of conventional electronic components."

4.4 Ground D — Claims 12–15 (off-grid "AC Master")

Combination: US 7,710,077 B2 (HDM, primary for the off-grid/load-management aspects) + US 2011/0012430 A1 (architecture) + US 2007/0040532 A1 (message system) + R. Jones, Measurement of Lumped Parameter Impedance (1974) for the load detector.

• Claim 12's distinguishing elements map as follows:
  • "load detector… arranged to detect the impedance of the connected AC load" → Jones (NPL), which is cited on the face of the '488 and expressly relied on by the '488 specification itself for exactly this circuit ("The load detector in this embodiment can be designed using standard LRC meter impedance measurement circuits and mechanism such as those described in the book…"). This is an applicant admission that the claim-12 load detector was known.
  • "check the impedance of the AC load to determine if it is within predetermined specifications," "continually determine whether the AC load is too large or too small," "turn the power off and trigger an error signal if the load is too large or too small" → HDM's current-sensing + programmable-threshold shutdown of the inverter output when load current falls below a minimum-loading limit ("In order to conserve energy… the inverter automatically shifts into the sleeping mode when output AC current is less than a predetermined threshold (e.g., 0.3 AAC)… The set point… is programmable"), and HDM's "the predetermined algorithm, the predetermined condition, and the predetermined period of time are user programmable." Over/under-load protection against programmable set points is standard inverter practice.
  • "initially energize the internal and external AC powerline" / "continually deliver AC power… to allow the other power inverters also connected on the same external powerline to synchronize" (the AC-Master / grid-forming role) → JP H09-275637 (Sanyo, "Fuel cell parallel operation system"), listed on the face, is the natural citation for parallel-run inverters establishing and sharing a common AC line, and US 2011/0273022 A1 (Dennis, "Method and Apparatus for Controlling a Hybrid Power System") for islanded/off-grid master-slave operation. Both unverified in this pass — flag as needing confirmation. The concept of a designated master inverter forming the voltage reference that others phase-lock to is well established in paralleled-inverter and microgrid practice; the '488 spec's own cross-reference to provisional 61/495,540 indicates the scheme was already described in the applicant's earlier off-grid filing (a §103(c)-disqualifiable reference — see §7).
  • "line sensing circuit… detect if there is AC power on the internal AC powerline prior to startup" → inverse of and complementary to '430's grid phase/zero-crossing sensing; HDM's shore-power/inverter changeover sensing is analogous.
  • Message system (claim 12(m), claims 13, 14) → same as Ground A: HDM's LED color + blink-cadence status panel is the single best reference here, because it is an off-grid inverter/charger with an LED-based status vocabulary and a remote panel.
• Claim 14 tracks Table 3 (including Case 13's "AC is present" and "Failed impedance test" and Case 14's "AC Load Error" — i.e., exactly the HDM/Jones concepts).

4.5 Ground E — Claims 16–17 (method of indicating status)

Claim 16 is a bare monitoring-plus-LED-actuation method with a five-state system-LED requirement. Its scope is broad and its §103 exposure highest:

Combination: US 2007/0040532 A1 (Bae) + US 7,710,077 B2 (HDM), in view of US 2011/0012430 A1.

• Bae: microcomputer determines inverter error/warning and displays it (LED among the display options) — i.e., "monitoring the status of a power inverter… and actuating LEDs based on the monitoring."
• HDM: LED color and pattern encode inverter operating states — the solid/blink vocabulary of Table 1.
• '430: supplies the monitored quantities (per-channel boost voltage, AC current/voltage, grid phase/zero-crossing) and the states (working / warning / error / off / grid-fault), so the content of the five indicated conditions is not inventive; only the mapping to a single LED is claimed.
• Claim 17 adds the per-channel five-state set, met as in §4.1 (claim 4) — '430's per-channel monitoring + Bae/HDM's indicator paradigm.

Claim 16's five enumerated states are the classic case of predictable mapping of already-computed internal state variables onto an already-known user interface. Absent objective evidence of unexpected results, this is the archetypal KSR combination.

5. Claim-by-claim summary of the §103 case

6. Why a PHOSITA would have combined these (consolidated motivations)

1. Same field, same problem, expressed in the art. Rotzoll states the exact problem (no outright indication of which module failed; only a production drop) and the exact market driver (field serviceability without taking the array offline).
2. '430's own cost rationale creates the multi-channel architecture and, with it, an express need for channel-level diagnosability.
3. HDM establishes the specific solution vocabulary (2-color/solid-vs-blinking LED status panels on an inverter with a remote panel) three and a half years before the '488 priority date.
4. Bae establishes that automatically detecting and displaying inverter error and error details is a known, desirable controller function — including displaying which fault type, via reference-value comparisons.
5. Choi establishes per-inverter addressing and controller-driven indicator displays across a population of inverters — the addressing/reporting behavior of claims 9–11.
6. The '488 specification's own admissions ("well known" converters/inverters per Rashid; "known combinations of conventional electronic components"; Jones for the load detector; MC4 as the industry standard) shrink the inventive space to the selection and arrangement of known elements.
7. KSR rationales available: (i) combining prior-art elements according to known methods to yield predictable results; (ii) simple substitution of one known element for another (LCD → LED; single status lamp → per-channel lamps); (iii) use of a known technique to improve a similar device in the same way (Bae's motor-drive fault display → PV inverter fault display); (iv) "obvious to try" with a finite number of identified, predictable solutions (color/pattern codes for five states); (v) design incentives/market pressure explicitly recited in both the prior art and the challenged spec (reduce installation and service cost).

7. The decisive caveat: pre-AIA §103(c) and common ownership

This is the single biggest weakness of the strongest grounds above, and any realistic challenge must confront it.

• US 2011/0012430 A1 published 2011-01-20, i.e., less than one year before the '488's 2011-06-30 priority date, so it is not §102(b) art. Its only §102 basis is §102(e) (as a published application, effective as prior art as of its 2009-07-16 filing date).
• The '488 specification identifies that same application as the work of "we" (Ser. No. 12/837,162) and shares two named inventors (Cheng, Mulkey) with the '488. US 2012/0212065 A1 (scalable and redundant mini-inverters, filed 2011-02-15) and US 2012/0313443 A1 (off-grid mini-inverters, priority 2011-06-10) are likewise §102(e)-only art of the same family/assignee lineage.
• Pre-AIA 35 U.S.C. § 103(c)(1) (as amended by the CREATE Act) provides that §102(e), (f), and (g) art shall not preclude patentability under §103 where "the subject matter disclosed and the claimed invention, not later than the time the claimed invention was made, were owned by the same person or subject to an obligation of assignment to the same person."
• Therefore, the architecturally near-identical '430 may be disqualified as §103 prior art, leaving a challenger to rebuild claim 1 from third-party art (Rotzoll '118 and '317, Kajouke, Pfeifer, Bhavaraju, Miasole, Marroquin, etc.) — a materially harder case, especially for the DC-combiner parallels and the combiner-fed bias supply.
• Practical consequences: (a) the disqualification is a factual defense requiring ownership evidence (assignment records, joint-research-agreement evidence) and is not self-proving from the face of the references; (b) even if '430 is disqualified as prior art, it remains powerful evidence of the level of ordinary skill and of the applicant's own characterization of the art, and the '488's incorporation-by-reference of it is an admission that the architecture was previously described by the same team; (c) if the challenger uses the '430 as a §102(e) reference for the architecture while relying on third-party art for the message system, the patentee will still argue the §103(c) disqualification taints the whole combination — the standard response is that the disqualification removes the reference from the combination, not the analysis, so a parallel all-third-party ground should be pleaded in the alternative.

A "clean" all-third-party alternative ground: Rotzoll '118 + Rotzoll '317 (module-mounted inverter, bracket, daisy-chain AC interconnect, PLC reporting of inverter and PV-module status and ID) + Bae + HDM, motivated by Rotzoll's own failure-identification problem — with multi-input combining supplied by any of the multiple-DC-source references on the face (e.g., RE37,126's separate DC sources; US 2004/0264225's per-source MPPT). This ground is weaker on the parallel-combiner and grid-sync elements but immune to §103(c).

8. Secondary considerations / objective evidence

• Expectation of success / no unexpected results: The '488 asserts benefits that are the inherent, predicted results of the combination — installer troubleshooting, reduced installation cost, usability with many channels. Under KSR, "the results of ordinary innovation… are not the subject of a patent," and the patent identifies no threshold showing or unexpected magnitude.
• Long-felt need: the patent frames installation cost as a large fraction of system cost; but Rotzoll identified the same need in 2007 and HDM delivered LED status panels on an inverter in 2006–2007, so the art was already moving to the solution — which undercuts a long-felt-need narrative.
• Commercial success / copying: the earlier sections note an active CyboEnergy litigation campaign (N.D. Cal., W.D. Tex., C.D. Cal., and matters involving Duracell, Home Depot, Yotta, Deye ESS, Generac). Accused-product success could be offered as objective evidence, but nexus is the weak point: any commercial success must be tied to the claimed LED message system / enclosure rather than to the underlying inverter performance (which the earlier sections note is the subject of sibling patents US 8,786,133 and US 9,331,489).
• Examiner allowance over 64 references of record: the patentee's strongest procedural argument is that Bae, Choi, HDM, Rotzoll, Pfeifer and Bhavaraju were all of record and the claims still issued. A successful challenge must therefore show something more than the references' presence — either an express reason to combine (Rotzoll's own problem statement is the best such evidence) or a claim construction that eliminates the perceived distinction (see §9).

9. Claim-construction leverage (cuts both ways)

Carried forward from the earlier section: "message system" in claims 1(m) and 12(m) is arguably functional. Under § 112(f) / pre-AIA ¶6:

• If construed as means-plus-function, it is limited to the disclosed structure — LED driver circuit 118/158 plus the FIG. 7/8 case-statement subroutines — and equivalents. That narrowing helps the patentee against broad prior art but hands the challenger a strong written-description/enablement-free obviousness argument, because the disclosed structure (a microcontroller-driven transistor current driver setting LED color and pattern) is precisely what Bae and HDM disclose.
• If construed as a nonce term with its ordinary meaning (a subsystem that conveys status), the claim is broader and correspondingly easier to invalidate over Bae/HDM.
  Either way, claims 10 and 11 — which expressly recite the case-number lookup and LED actuation subroutine — should be construed narrowly to the described case tables, and then assessed for obviousness as the routine implementation of a lookup table mapping fault codes to indicator states (a characterization Bae's and Choi's controller-driven displays support).

10. Bottom line

1. Claims 16–17 are the most vulnerable, followed closely by claims 9–11. Claim 16 is a broad monitoring-plus-LED-actuation method whose five states are the already-existing internal state variables of the '430 architecture; Bae and HDM together supply the "detect inverter fault → drive LED status display with color/blink coding" step. Claim 9's steps (a)–(c) and (e)–(g) appear verbatim in '430's claim 24.
2. Claims 1–5 are obvious over US 2011/0012430 A1 in view of Bae (US 2007/0040532 A1) and HDM (US 7,710,077 B2), with the express motivation supplied by Rotzoll's US 2008/0283118 A1 ("no outright indication that a PV module has failed"). Rotzoll is the linchpin of the motivation argument and is squarely analogous art.
3. Claim 12 (off-grid) is obvious over HDM (off-grid inverter/charger with programmable minimum-load shutdown and watertight enclosure) in view of '430, with the impedance-based load detector supplied by R. Jones (1974) — cited on the face and confessed in the specification — plus parallel-inverter/AC-master art (JP H09-275637 and/or US 2011/0273022 A1, both unverified here).
4. Claims 6–8 and 15 (the enclosure) are the weakest §103 target, though Rotzoll (module-back bracket, daisy-chain AC interconnect, metal enclosure) plus HDM (extruded aluminum watertight case, glands for AC input/output cables, frame mounting) supplies every enumerated part; the remaining question is whether the specific arrangement would have been obvious, and the '488's own admission that MC4 connectors are the industry standard and that alodine/aluminum is routine outdoor practice substantially weakens any nonobviousness argument.
5. The case is materially weaker if US 2011/0012430 A1 (and the other Cheng/CyboEnergy §102(e) references) are disqualified under pre-AIA § 103(c). Any serious challenge should therefore plead (i) the primary ground relying on those references, and (ii) an alternative all-third-party ground anchored on Rotzoll '118 + '317 + Bae + HDM, so that a common-ownership ruling does not collapse the entire invalidity case.

Items I could not verify and would not represent as established: the specific disclosures of JP H09-275637, JP 2001-008383, RE37,126, US 2010/0202177, US 2004/0264225, US 2008/0111517, US 2009/0160258, US 2009/0174259, US 2012/0104872, US 2009/0000654, and US 2011/0273022. These appear on the '488's citation list (per the Justia citation table at https://patents.justia.com/patent/9331488 and the Google Patents record), and I have assigned them roles based on title/assignee only. Each should be checked against its specification before being relied on in a written §103 analysis.