US 11326940

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

US 11,326,940 B2 — Summary

Identifier interpreted literally: "11326940" → US Patent 11,326,940 B2 (application 17/032,441). I did not substitute any similar number.

Bibliographic data (from the patent record)

Abstract (as issued)

An ambient light signal adjustment method, a chip and electronic equipment. The method includes receiving a first ambient light signal which is an ambient light signal acquired by a photo plethysmor graph (PPG) sensor; then determining, according to an intensity of the first ambient light signal automatically, whether to turn on an ambient light cancellation circuit which is configured to adjust the first ambient light signal according to its intensity; and if it is determined to turn on the ambient light cancellation circuit, adjusting a cancellation signal intensity of the circuit automatically to enable an intensity of the adjusted first ambient light signal to be within a preset range. "Adjustment of the ambient light signal acquired by the PPG sensor is realized, thereby ensuring accuracy of a wearable device to detect a physiological characteristic of a user."

Independent claims in plain language

• Claim 1 (method). Three steps: (1) receive a "first ambient light signal" that is ambient light acquired and output by a PPG sensor; (2) automatically decide, based on how intense that signal is, whether to switch on an ambient light cancellation circuit — the circuit being one configured to adjust the first ambient light signal according to its intensity; (3) if the decision is to turn it on, automatically set the circuit's cancellation signal intensity according to the first ambient light signal's intensity so that the adjusted first ambient light signal falls within a preset range.
• Claim 8 (chip). A chip with two cooperating blocks: a determining circuit wired to an ambient light cancellation circuit. The determining circuit receives the first ambient light signal from the PPG sensor and decides whether to turn the cancellation circuit on based on that signal's intensity; the cancellation circuit adjusts the signal per its intensity; and if turned on, it automatically adjusts its cancellation signal intensity so the adjusted first ambient light signal intensity lands within a preset range.
• Claim 20 (electronic equipment). Electronic equipment comprising the chip of claim 8 (a bare apparatus claim to the host device).

Dependent claims add: the >first-preset-threshold turn-on test (claim 2) / ≤threshold stay-off test (claim 3); determining the electrical signal intensity and dynamically driving the adjusted signal to ≤a second preset threshold to yield a "second ambient light signal," with cancellation intensity ≤ electrical signal intensity (claim 4); subtracting the first or second ambient light signal from a received "mixed light signal" (target light + ambient) to recover the target light signal (claim 5, and chip claims 12/13); proportionality of cancellation intensity to ambient light intensity (claim 6); a successive approximation loop plus digitally controlled current source (claim 7); comparator as the determining circuit (claim 14); amplifier/driver (16); ADC/DAC (17); PPG sensor light-emitting device and photoelectric conversion device, with the photoelectric conversion device on-chip (18) and the light-emitting device integrated in the chip (19).

Observations / drafting quirks worth noting

• Claim 3 depends on claim 1 but recites "the first preset threshold," a term first introduced in claim 2 — a potential antecedent-basis (35 U.S.C. § 112(b)) vulnerability. Claim 4 likewise depends only on claim 2, so it captures only the "greater than" branch.
• The granted claim 1 recites "an intensity of the adjusted first ambient light signal," whereas the summary-of-invention text and the original pre-grant publication used "first light signal" in that clause — a minor wording divergence between prosecution text and granted claim.
• Both claims 12 and 13 depend on claim 11 and split the two processor alternatives (recovering the target signal via the first vs. the second ambient light signal) rather than using a Markush-style "or."

Litigation / docket check (important caveat)

I ran searches aimed at USPTO records and CAFC 2026 dockets for this patent number. No CAFC appeal, district-court case, or PTAB/IPR proceeding involving US 11,326,940 surfaced in the results. I found no evidence that this patent is in litigation as of the search date. However, I must be explicit about uncertainty: my searches returned general patent-office and Federal Circuit material but did not surface an authoritative live docket entry for this patent, and I do not have direct query access to USPTO Patent Center, the assignment database, or PACER/CAFC CM-ECF. Therefore I cannot affirmatively represent that no such proceeding exists — only that none appeared in my search results.

For contrast, the record does show substantial prior-art citation activity (e.g., Nellcor's US 2005/0187448 A1 "Oximeter ambient light cancellation," Fitbit's US 2014/0275852 A1 and US 2014/0288435 A1, TI's US 9,615,427 B1 and US 2019/0175037 A1, Nonin's US 2019/0298237 A1, and Goodix's own CN109923785A "Current cancellation circuit, heartbeat detection device and wearable device"), plus a non-patent citation to Jongpal Kim et al., "Ambient Light Cancellation in Photoplethysmogram Application Using Alternating Sampling and Charge…" (title truncated in the record).

If you need certified litigation status, the authoritative next steps would be a PACER/CM-ECF docket search, the CAFC's online docket, and USPTO Patent Center's "Patent Litigation" and assignment tabs — none of which I could query directly here.

I'll search for litigation involving US patent 11326940.

The initial searches returned irrelevant results. Let me try more targeted queries.

No litigation results yet. Let me try more specific litigation databases and query forms.

Summary of Findings

No litigation involving U.S. Patent No. 11,326,940 ("US11326940") was identified in the searches I performed.

What I searched

• General web queries for the patent number ("11326940"), the formatted number ("11,326,940"), and the "US11326940" identifier in combination with litigation, lawsuit, docket, and infringement terms.
• Patent-owner-focused queries (Shenzhen Goodix Technology Co., Ltd.) tied to infringement/heart-rate/wearable litigation.
• Sources indexed in the results included court documents (CourtListener/RECAP), PTAB filings, RPX Insight litigation documents, and the Google Patents page for the patent itself.

Results

Every hit that referenced a "640 patent" was a false positive — those dockets concerned unrelated patents (e.g., an Impax/Qilu ANDA pharmaceutical patent and the Pacira '940 patent, U.S. Pat. No. 12,156,940). None referenced U.S. Patent No. 11,326,940, which is the Goodix PPG ambient-light-cancellation patent.

The Google Patents record for US11326940 (fetched 2026-09-24) itself contains no "Litigation" section, no asserted-patent docket listings, and no PTAB proceedings. It lists:

• Assignee: Shenzhen Goodix Technology Co., Ltd.
• Inventor: Chang Liu
• Priority: 2019-10-25 (CN 201911021430.0)
• Filed: 2020-09-25; Granted: 2022-05-10; Adjusted expiration: 2041-01-30
• Family members: EP3811860A1, KR102412279B1, CN110604559B, WO2021077807A1

A patent that had been asserted would typically show up in these records, in PTAB post-grant proceedings, or in court dockets — and none did.

Important caveats

• I did not obtain direct live access to the Unified Patents litigation portal, PACER, or Docket Navigator's full datasets (their search interfaces require direct database queries/login), so I cannot rule out a very recently filed and not-yet-indexed case.
• The absence of indexed results is strong but not conclusive evidence of no litigation. It is also worth noting that U.S. patents in this space are often asserted in the Eastern District of Texas, District of Delaware, or W.D. Texas; a targeted Docket Navigator / Unified Patents / Lex Machina query on "11,326,940" and on the Goodix patent family would be the appropriate next step for confirmation.

Conclusion: Based on the available search results, there is no known litigation involving U.S. Patent No. 11,326,940 as of the current date. If you need a definitive negative, I'd recommend a direct query to Docket Navigator or PACER for the exact patent number.

I'll verify the structured "no proceedings" signal against live sources before finalizing.

Let me run a few more targeted checks before concluding.

Proceedings overview

Total AIA trial proceedings on file: 0 (zero — no IPR, no PGR, no CBM). The structured "PTAB proceedings on file" block reports that the USPTO Open Data Portal returns no AIA trial proceedings for US 11326940, and my independent web searches (including queries for IPR numbers tied to the patent number, to assignee Shenzhen Goodix, and to the '940 family) surfaced no petition, institution decision, FWD, or Federal Circuit appeal naming this patent. Because there is nothing invalidated and nothing sustained in a trial, the defensive posture is neutral-to-unhelpful for a defendant: US 11326940 is an untested, fully intact patent — all 20 claims stand as issued, with no PTAB estoppel, no claim-cancellation certificate, and no PTAB/CAFC record to point to. Any IPR defense would be a first-instance, from-scratch attack on a patent that has never been through the Board.

I found no proceeding to report on, so the per-proceeding sections below are used to document the negative findings and the adjacent signals that matter.

No proceedings identified — negative finding detail

• Type: N/A (no Inter Partes Review, Post-Grant Review, or Covered Business Method review on file)
• Filed: N/A
• Status: No PTAB activity as of the USPTO ODP ingest reflected in this prompt (checked 2026-09-28 against live web sources; none found)
• Judge panel: N/A — no panel empaneled
• Petition grounds: N/A — no petition
• Institution decision: N/A
• Final Written Decision: N/A — no claim of US 11326940 has ever been canceled or confirmed by the Board
• Settlement / termination: N/A
• Appeal: N/A — no FWD, therefore no CAFC appeal of an FWD. The patent itself is not the subject of any Federal Circuit opinion I could locate.
• Defensive value: Zero PTAB precedent cuts both ways. You cannot rely on an existing cancellation to knock out a demand letter grounded in claims 1–20, but you also face no adverse institution-denial record, no patent-owner-favorable FWD, and no § 315(e) estoppel ratchet binding anyone else. The Board's docket on this patent is a blank slate.

Strategic summary

Claim status: all 20 claims are UNTESTED and intact. Claims 1–7 (method), 8–19 (chip), and claim 20 (electronic equipment comprising the chip of claim 8) are all live, unamended, and uncontested at the PTAB. Nothing has been canceled, so there is no "safe" subset of claim numbers you can simply avoid; conversely, there is no FWD finding of validity that the patent owner could brandish. The sole independent method claim is claim 1; claim 8 is the sole independent chip claim; claim 20 is a one-line apparatus claim that piggybacks on claim 8 — the classic wide-net drafting that makes even a narrow product (a PPG front-end chip with an ambient-light cancellation circuit) a plausible target.

Estoppel landscape: § 315(e)(2) estoppel is not yet in play. Because no IPR has been instituted, no petitioner (or privy) is estopped from raising any ground, and the patent owner has not yet been able to argue that any art was "raised or reasonably could have been raised." For a defendant being asserted against today, the § 315(b) clock is the binding constraint: you must file within one year of service of the complaint alleging infringement. The PGR window, by contrast, is closed — PGR petitions must be filed within nine months of grant under 35 U.S.C. § 321(c), and the patent granted 2022-05-10, putting the PGR deadline at approximately 2023-02-10. CBM is unavailable on two independent grounds: the transitional program sunsets for petitions filed on or after 2020-09-16, and a PPG ambient-light cancellation circuit is not a "covered business method" patent. That leaves IPR as the only AIA vehicle — § 102 / § 103 art only, no § 112 or § 101 theories.

Pattern signals. No serial petitioner, no defensive aggregator. I searched specifically for Unified Patents or any other aggregator challenging the Goodix PPG portfolio and found nothing tying them to US 11326940. The family is otherwise active rather than embattled: CN110604559B (the Chinese parent, priority 2019-10-25), KR102412279B1, and EP3811860A1 all remain in force/active, while the PCT member WO2021077807A1 is recorded as ceased — a prosecution-stage lapse of the international application, not a validity loss that has any estoppel or preclusive effect in the US. One adjacent item worth flagging, because it is the closest thing to a third-party validity attack I found: the Extended European Search Report dated 2021-02-11 on EP 20199848.1 (the EP counterpart, EP3811860A1) applied CN 109 863 703 A — a Shenzhen Goodix application titled "Signal conditioning method, device, chip, device and storage medium" — as category-X art against claims 1–15, i.e., the EPO examiner treated the applicant's own earlier CN filing as destroying novelty of the counterpart's claim set. That reference is also cited on the '940 face, along with CN109923785A ("Current cancellation circuit, heartbeat detection device and wearable device") and the Fitbit heart-rate-monitor family (US 2014/0275852, US 2014/0288435, US 9,662,053). This is EPO/EU activity only; it is not a PTAB proceeding, creates no US estoppel, and the EP application remains pending rather than revoked — but it is a signal about where a US petitioner's art-search should start.

Recommended next steps

• No PTAB proceeding exists — treat that as the headline. Do not cite an IPR number in a litigation brief or a demand-letter response; none exists. Verify directly at PTAB E2E/Case Search (https://ptacts.uspto.gov/ptabweb/) by searching patent number 11326940, and confirm the ODP record before relying on this memorandum.
• If you have been served with a complaint asserting US 11326940, the IPR clock is running. Under § 315(b) you have one year from service to petition. Also screen for § 315(a)(1): if your client filed a declaratory-judgment action of invalidity first, IPR is barred.
• Grounds must be § 102/§ 103 only. Prioritize the three Goodix-family prior filings on the face of the patent plus the Fitbit family, and mirror the CN109863703A theory the EPO's search division already adopted against the EP counterpart's claims 1–15 — an X-category usage by a national examining division is useful corroboration for a US obviousness/anticipation theory, though it is not binding on the Board.
• No claim has been canceled, so an IPR-based defense requires winning, not just filing. Budget for a full trial: institution decision at ~6 months from petition, and a Final Written Decision within 1 year of institution under § 318(a) (extendable to 18 months for good cause under § 316(a)(11)).
• Watch the Chinese parent. CN110604559B and EP3811860A1 are the family members most likely to attract a validity challenge first (CNIPA invalidation, EPO opposition). A foreign cancellation of the parent would not bind the PTAB on the '940, but it would materially sharpen a US obviousness narrative. No such proceeding is currently on file to my knowledge — I could not confirm any, and I will not speculate.

Confidence note: I have high confidence in the negative PTAB finding (structured ODP data plus four independent web searches found nothing), and moderate confidence in the EP search-report detail, which comes from the EPO publication-server copy of the EP3811860A1 search report. I did not find any Federal Circuit docket entry, settlement, or litigation naming US 11326940; if one exists, it is not yet indexed in the sources I can reach.

I'll run targeted searches to verify the assignment record and check for NPE signals.

Let me dig for the specific assignment reel/frame and correspondent details, plus Goodix corporate status.

I hit the search limit while gathering assignment-center and corporate details, so I'll build this from the authoritative patent record (which contains the Google Patents legal-events/assignment data) and the litigation/filing material I did retrieve, and flag explicitly what I could not verify.

---

Inventors

No inventor-departure pattern is detectable from the records reviewed: the patent has a single inventor, and he remains associated with Goodix on contemporaneous filings. There is no evidence of inventors leaving the assignee within 12 months of filing, which would have been a fire-sale precursor.

Original assignee

Shenzhen Goodix Technology Co., Ltd. — the entity named on the issued patent (original and current assignee).

• Line of business: Fabless semiconductor designer of fingerprint sensors, touch controllers, and optical/PPG biosensing chips for smartphones and wearables. This patent (ambient-light cancellation for a PPG/heart-rate sensor front-end) sits squarely in Goodix's sensor-chip product line.
• Product embodying the claims: Goodix sells heart-rate / PPG sensor chips and optical sensor chips; the claims are directed to the chip architecture for those products. I did not retrieve a specific datasheet tying a part number to these claims, so this is corroborated but not document-linked.
• Public-company status: Goodix is a publicly listed company (Shanghai Stock Exchange; I recall ticker 603160, though I could not re-verify the ticker this session). It is an active operating company, not a licensing vehicle.
• Litigation posture (context, not this patent): Goodix is a frequent participant in patent disputes as both defendant and plaintiff — e.g., it was sued by Synaptics (N.D. Cal. + ITC §337, 2015, settled 2016) and by Fingerprint Cards (Beijing IP Court, 2018), and it has asserted its own patents against competitor Silead Microelectronics (fingerprint-recognition patent ZL201410204545.4). This confirms an operating-company, competitor-facing enforcement posture, but no suit naming US 11,326,940 was found.
• Current status: Operating.

Assignment timeline

The Google Patents legal-events record for US 11,326,940 shows one recorded assignment entry. I could not retrieve the reel/frame number or the correspondent of record — the sources available to me (Google Patents legal events) expose the conveyance, dates, and parties but not the reel/frame or the recording attorney/agent. I am not fabricating those fields.

• Executed date: not stated in the record / recorded 2020-09-25 — Reel/Frame not retrievable from available sources
  • Conveyance: Assignment — "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)"
  • Assignor: Liu, Chang (the sole inventor)
  • Assignee: Shenzhen GOODIX Technology Co., Ltd.
  • Correspondent: not available in the records I could query (USPTO Assignment Center reel/frame and correspondent field were not reachable this session). Cannot assess repeat-correspondent recurrence.
  • Context: Routine inventor-to-employer assignment of a first-filed application — captured the same day the US application (17/032,441) was filed, ~11 months after the CN priority filing (CN 201911021430.0, 2019-10-25). This is a standard corporate ownership record, not an acquisition, fire-sale, securitization, or transfer-to-asserter.
• No post-issuance assignment is recorded. The patent issued 2022-05-10 (US 11,326,940 B2) and, per the record, remains with Shenzhen Goodix Technology Co., Ltd. No security interest, merger, change-of-name, license, or release is recorded against this patent.

Because the Assignment Center itself returned no additional entries beyond the original assignment, the practical reading is that the original assignee still owns the patent — which is itself the finding here.

Timeline diagram

timeline
    title Ownership of US 11326940
    2019 : CN priority filed by Goodix
    2020 : US application filed
         : Inventor assigns to Shenzhen Goodix
    2022 : US patent issued to Goodix
         : No further assignments recorded

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The only transfer is inventor → operating company (Shenzhen Goodix, recorded 2020-09-25). No "IP/Holdings/Licensing/Ventures" entity appears anywhere in the chain; the assignee is a public operating semiconductor company, not a single-purpose LLC at a registered-agent address.

2. Known asserter in the chain — not present. Neither assignor nor assignee matches any entity on the cited NPE lists (Acacia, Marathon, IV, Wi-LAN/Conversant, Vringo, Pendrell, Round Rock, etc.). Current assignee is Goodix, an operating chip vendor.

3. Repeat correspondent across the chain — unclear / cannot determine. Only one assignment exists, and I could not retrieve the correspondent of record (the field was not exposed in the reachable sources). A single appearance would not be a "repeat" signal in any event, and no recurrence can be shown. Flagged as a verification gap.

4. Cascading transfers — not present. Zero consecutive post-filing transfers; no chained LLCs, no shared-address cluster, no sub-24-month relay.

5. Pre-litigation transfer — not present. No assignment within 6 months before any suit; no suit on this patent was found at all. The only assignment coincides with the filing date and is a first-instance employer assignment, not a standing/venue-setup transfer.

6. Bankruptcy fire-sale — not present. No Chapter 7/11, no reorganized-assignee, no judicial sale associated with Goodix or this patent.

7. Privateering — not present. No transfer of this patent from Goodix to an NPE asserting on Goodix's behalf. (Goodix asserts its own patents directly as a competitor, per the Silead matter — that is ordinary operating-company enforcement, not privateering.)

8. Defensive aggregator (anti-NPE) — not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN; it terminates at the original operating company.

Verdict

Insufficient data — under the given taxonomy, a chain consisting solely of the original assignment maps to this category. That said, the record is affirmatively clean, not merely thin: the single recorded entry is the inventor→Shenzhen Goodix assignment dated/recorded 2020-09-25 (Google Patents legal events), there are no post-issuance transfers, no security interests, and no shell/licensing entity anywhere — and the assignee is a publicly listed operating semiconductor vendor that litigates as a competitor. No NPE/troll signal is present; the patent simply remained with its original, operating-company owner.

Two verification gaps I could not close this session, stated plainly: (a) the reel/frame number, and (b) the correspondent of record for the 2020-09-25 assignment. Both require the USPTO Assignment Center record, which I could not query directly.

Verify at: https://assignmentcenter.uspto.gov/ (search patent number 11326940) and the mirrored index at https://assignment.uspto.gov/patent/index.html.

I'll verify the key prior-art references cited on the face of US 11,326,940. Let me run several targeted searches in parallel.

Let me verify the two remaining references I haven't confirmed — the Goodix CN109863703A and the Seiko Epson JP reference.

Prior-Art Analysis — U.S. Patent No. 11,326,940 B2

"Ambient light signal adjustment method, chip and electronic equipment"

Patent under review interpreted literally as US 11,326,940 B2 (application 17/032,441; priority 2019-10-25 via CN 201911021430.0; granted 2022-05-10; assignee Shenzhen Goodix Technology Co., Ltd.). Searches were run against the specific number 11326940; no similar-number results were substituted.

This section builds on — and does not repeat — the earlier-generated claim set (20 claims; independents 1, 8, 20), the obviousness combinations, the litigation/PTAB negative findings, and the assignment history. Where my findings refine an earlier statement, I flag it explicitly.

---

A. The § 102 framework applied here

A claim is anticipated under 35 U.S.C. § 102 only if a single prior-art reference discloses each and every limitation, arranged as in the claim. Because the '940 patent is an AIA patent (effective filing 2019-10-25), the relevant categories are § 102(a)(1) (patented, described in a printed publication, or in public use/on sale before the effective filing date) and § 102(a)(2) (described in a U.S. patent/application that names another inventor and was effectively filed first).

The critical consequence for this patent: every independent claim (1, 8, 20) requires an "ambient light cancellation circuit" that is (i) conditionally turned on based on the ambient signal's intensity and (ii) whose "cancellation signal intensity" is automatically adjusted to bring the adjusted ambient signal within a preset range. No single cited reference discloses that complete combination. As a legal matter, none of the 16 prior-art patent citations anticipates claims 1, 8, or 20, and therefore none anticipates the dependent claims either (a dependent claim cannot be anticipated unless its base claim is). Below I identify, for each reference, the near-miss limitations and whether any narrower claim is arguably met — and I state plainly where the answer is "none."

Dating note (all references verified to predate the 2019-10-25 priority date, except the family self-citation):

---

B. Reference-by-reference analysis (the citation list on the face of US 11,326,940)

1. US 5,644,385 A — Hamamatsu Photonics K.K.

"Distance measuring device using position sensitive light detector" — priority 1994-04-28; granted 1997-07-01. Inventor/assignee: Hamamatsu Photonics K.K.
Description (verified): A triangulation distance meter using a position-sensitive detector (PSD). The signal-extraction circuit 20 contains a "noise cancellation circuit" that cancels background-light and noise components from the photocurrents (it pre-integrates a background value and subtracts it, using MOS transistors and constant-current sources).
§ 102 assessment: None. This discloses background-light cancellation in an optical-sensing front end, but there is no PPG sensor, no conditional on/off decision keyed to ambient intensity, and no automatically varied cancellation current. It is at most generic § 103 corroboration that background cancellation in photodiode front-ends was old.
Source: https://patents.google.com/patent/[US5644385A](/patent/US5644385A)/en

2. US 2005/0187448 A1 — Nellcor Puritan Bennett Inc.

"Oximeter ambient light cancellation" — filed 2004-02-25; published 2005-08-25.
Description (verified): A pulse oximeter that measures the "dark"/ambient level before and after each LED period, averages it, and subtracts it from the detected signal; preloads filters when gain/LED power changes.
§ 102 assessment: None (near-miss on claim 5). It discloses sampling ambient light during dark periods from an optical physiological sensor — the "receiving a first ambient light signal … acquired and output by a PPG sensor" element of claim 1 — and the mixed-signal-minus-ambient subtraction underpinning claim 5. But it recites no "ambient light cancellation circuit" that is conditionally turned on or whose cancellation signal intensity is automatically adjusted. Claim 5 depends on claim 4→2→1, so no complete anticipation. Strong § 103 art.

3. US 2019/0298237 A1 — Nonin Medical, Inc.

"LED control utilizing ambient light or signal quality" (Isaacson, Schilling; app. 16/292,024 filed 2019-03-04; earliest priority 2008-06-03) — published 2019-10-03.
Description (verified): Controls LED drive current and/or integration time as a function of detected ambient light or signal quality. It compares an Ambient-Light measurement to a No-Light measurement and to IR/Red DC measurements against thresholds, then boosts or reduces LED drive current; includes an "ambient light current amp" for ambient compensation.
§ 102 assessment: None (near-miss on claim 1's decision step). It discloses ambient-light measurement from an optical sensor and an automatic, threshold-based control decision responsive to that measurement — close to the "determining … automatically, whether to turn on" element. But what is adjusted is the LED drive current/integration time, not the intensity of an "ambient light cancellation circuit." Different quantity, so no anticipation. Published 2019-10-03 (≤ 1 year pre-priority) by a different inventor, so the § 102(b)(1)(A) grace-period exception does not remove it. Strong § 103 art.
Source: https://patents.justia.com/patent/[11246515](/patent/11246515)

4. CN 101944897 A — Intersil (Lin Xijian)

"Proximity sensors with improved ambient light rejection" (US 8,222,591 B2) — priority 2009-07-07; published 2011-01-12.
Description (verified): A photodiode proximity sensor using an ADC with analog-to-digital-to-analog (ADA) feedback: an M-bit ADC digitizes the ambient-only level; an M-bit DAC reconverts it to an analog current I_dac that is subtracted from the later signal; an N-bit ADC digitizes the residue.
§ 102 assessment: None (structural near-miss on claims 7/14/15). The M-bit-ADC / M-bit-DAC / subtractor feedback loop is functionally "a successive approximation loop and a digitally controlled current source" (claim 7) with a comparator (claim 14). But claims 7 and 14 depend on claims 1/8, and Intersil is a proximity sensor, not a PPG sensor, and does not recite the conditional turn-on. No complete anticipation; the transfer to PPG is a predictable application of the same photodiode front-end (§ 103).
Source: https://patents.google.com/patent/CN101944897A/zh

5. US 2014/0275852 A1 — Fitbit, Inc.

"Wearable heart rate monitor" — priority 2012-06-22; published 2014-09-18.
Description (record-based): Wearable PPG heart-rate monitor with ambient-light handling and adjustable acquisition.
§ 102 assessment: None. Discloses the wearable PPG context and ambient handling only. (Full text not independently re-verified this session — see Caveats.)

6. US 2014/0288435 A1 — Fitbit, Inc.

"Heart rate data collection" — priority 2012-06-22; published 2014-09-25.
Description (record-based): Heart-rate data collection with ambient-light handling.
§ 102 assessment: None. Same reasoning as ref. 5; § 103 corroboration for the wearable-PPG field.

7. US 9,662,053 B2 — Fitbit, Inc.

"Physiological data collection" — priority 2012-06-22; granted 2017-05-30.
Description (record-based): Physiological data collection in a wearable.
§ 102 assessment: None. Field/context art; no cancellation-circuit limitations.

8. CN 105852841 A — Fitbit (飞比特公司)

"Heart rate data collection" — priority 2013-06-03; published 2016-08-17.
Description (record-based): Chinese counterpart of the Fitbit heart-rate-collection family.
§ 102 assessment: None. Field/context art only.

9. CN 105491943 A — Royal Philips (皇家飞利浦有限公司)

"Photoplethysmography sensor apparatus and method" — priority 2014-06-30; published 2016-04-13.
Description (record-based): PPG sensor apparatus/method.
§ 102 assessment: None. At most relevant to claim 18's "PPG sensor comprises a light-emitting device and a photoelectric conversion device," but claim 18 depends on claim 8 and the reference is not shown to disclose the cancellation circuit. (Full text not re-verified — see Caveats.)

10. WO 2016/011173 A1 — Google Inc. (Mirov, Murphy)

"Context discrimination using ambient light signal" (granted as US 10,656,009 B2, Verily Life Sciences) — filed 2014-07-16; published 2016-01-21.
Description (verified): Measures ambient-light intensity with a light-sensitive element; a controller obtains a set of ambient-light measurements and determines whether they match a stored ambient-light profile corresponding to a particular environment/context.
§ 102 assessment: None. Discloses ambient-light sensing plus an automatic controller-based determination, but nothing about a PPG cancellation circuit, conditional turn-on, or adjustable cancellation intensity. § 103 corroboration for ambient-light sensing/classification only.
Source: https://patents.google.com/patent/WO2016011173A1/en

11. CN 106559095 A — Apple Inc. (苹果公司)

"Electronic equipment and correlation technique including ambient light compensation circuit that heart rate is generated" — published 2017-04-05.
Description (record-based, title/metadata only): Electronic equipment including an ambient-light compensation circuit used in heart-rate generation.
§ 102 assessment: None (unverified — flagged). Potentially relevant to combining ambient-light compensation with heart-rate sensing, but I was unable to fetch the full text this session; characterization is limited to what the title supports.

12. US 9,615,427 B1 — Texas Instruments Inc. (Narayanan et al.)

"Exploiting constructive interference from ambient conditions" — granted 2017-04-04.
Description (verified): An optical system receives a combined optical signal (ambient + interrogation component) and a separately measured ambient light signal. A correlation determination circuit computes a correlation factor; an ambient condition control circuit compares the factor to a low and high threshold, and then: (a) deactivates the LED if the factor exceeds the high threshold; (b) cancels the ambient component by subtraction if the factor is between the thresholds; or (c) leaves the combined signal alone if below the low threshold. The rationale given is expressly that subtracting ambient adds noise and worsens SNR, so cancellation is applied only when warranted.
§ 102 assessment: None on claims 1/8/20 — but the most interesting near-miss on the decision step. This reference discloses (i) receiving an ambient light signal, (ii) a circuit that automatically compares to thresholds and thereby decides whether to perform ambient cancellation, and (iii) ambient cancellation by subtraction. That maps strikingly onto claim 1's "determining … automatically, whether to turn on an ambient light cancellation circuit" — and TI even articulates the same noise-based motivation the '940 patent asserts. What TI lacks is an "ambient light cancellation circuit" whose cancellation signal intensity is automatically adjusted according to the intensity of the ambient signal: TI's decision is keyed to correlation, and cancellation is a binary subtraction. No complete anticipation. Ranked as a top § 103 reference, and the single best teaching that conditional cancellation was known.
Source: https://patents.google.com/patent/US9615427

13. US 2019/0175037 A1 — Texas Instruments Inc.

"Motion detection and cancellation using ambient light" — priority 2017-12-08; published 2019-06-13.
Description (verified): Wearable heart-rate monitor: LED on → reflected LED light + ambient; LED off → ambient only; subtract to isolate the LED signal; an adaptive filter whose coefficients are dynamically adjusted.
§ 102 assessment: None (near-miss on claims 1 and 5). Discloses LED-on/off subtraction (claim 5) and automatic, signal-dependent dynamic adjustment (claim 1's "automatically"). No ambient-light cancellation circuit with adjustable cancellation intensity conditionally turned on. § 103 corroboration that dynamic, intensity-responsive adjustment in wearable PPG was routine by 2017–2019.

14. JP 2019-134089 A — Seiko Epson (セイコーエプソン株式会社)

"Photoelectric sensor, photoelectric sensor module, and organism information measurement device" — published 2019-08-08.
Description (record-based, title/metadata only): Photoelectric sensor / module for biological-information measurement.
§ 102 assessment: None (unverified — flagged). Potentially relevant to PPG hardware; not full-text verified this session.

15. CN 109863703 A — Shenzhen Goodix Technology Co., Ltd.

"Signal conditioning method, device, chip, device and storage medium" (信号调节方法、装置、芯片、设备及存储介质) — filed 2019-01-24; published 2019-06-07. Inventor: 刘畅 (Liu Chang) — apparently the same inventor as the '940 patent.
Description (verified abstract): Acquire a first background light signal; if its intensity is greater than a first preset threshold, automatically adjust the amplifier gain so that the adjusted first background-light-signal intensity is less than or equal to the first preset threshold, obtaining a second background light signal; acquire a mixed light signal; derive a target light signal from the mixed light signal and the first or second background light signal; if the target light signal intensity is less than a second preset threshold, automatically adjust the drive current of the light transmitter.
§ 102 assessment: This is the closest textual match to the '940's threshold-driven, automatic, subtraction-based scheme — but it does not anticipate the independent claims, and it is likely excepted from § 102(a) altogether. Two independent reasons:

• (a) Missing limitation. CN 109863703 A adjusts amplifier gain (and light-transmitter drive current), not "a cancellation signal intensity of an ambient light cancellation circuit," and it does not switch such a circuit on/off. The "ambient light cancellation circuit" limitation of claims 1/8 is absent. Its threshold-automatic-adjustment logic and mixed-signal-minus-background subtraction map onto elements of claims 1 and 5, but not the full claims.
• (b) Likely § 102(b)(1)(A) exception. The reference names the same inventor (Liu Chang) and published 2019-06-07 — within one year before the '940 effective filing date (2019-10-25). A disclosure made by the inventor within that grace period is excepted from § 102(a) prior art under § 102(b)(1)(A). It is also not § 102(a)(2) art, because § 102(a)(2) requires naming another inventor. Takeaway: this reference may be removable as prior art entirely — a refinement of the earlier sections, which analyzed the inventor-match point only for CN 109923785 A (a different inventor, Li Bo). Confirm the identity of 刘畅 ≡ Chang Liu before relying on the exception.
  Source: http://agri.nais.net.cn/patentdetails/FF87E7C0-E8B5-4733-BFD0-365A98718586.html (CN 201980000174.9)

16. CN 109923785 A — Shenzhen Goodix Technology Co., Ltd. (EP 3,754,854 B1; US 11,350,836 B2)

"Current cancellation circuit, heart rate detection device and wearable device" — PCT filed 2019-01-31; published 2019-06-21.
Description (verified): Two-stage PPG operation — LED off → interference/background current alone; a SAR ADC + comparator + DAC successively approximate and output an analog cancellation current; LED on → mixed current minus cancellation current. Background states that a background-light-cancellation (BGC) circuit keeps the channel operable but "the use of BGC will increase noise of the channel" and requires an algorithm.
§ 102 assessment: None on claims 1/8/20 — the closest structural patent reference for claims 7/14/15. It discloses the successive-approximation loop with a digitally controlled current source (claim 7), a comparator-based determining circuit (claim 14), and mixed-current-minus-cancellation recovery (claim 5 elements). But its cancellation is part of a staged operation, not an intensity-keyed conditional turn-on; the "determining … whether to turn on" limitation is not disclosed. Because claims 7/14/15 depend on claims 1/8, no complete anticipation exists — but these claims are anticipatory-in-substance on this element. Top-tier § 103 primary reference (consistent with the earlier obviousness section), and, being a printed publication by a different inventor (Li Bo), it is not saved by either the § 102(b)(1)(A) inventor exception or the § 102(b)(2)(C) common-ownership exception.
Source: https://patents.justia.com/patent/[11350836](/patent/11350836)

17. CN 110604559 A — Shenzhen Goodix Technology Co., Ltd.

"Ambient light signal adjusting method, chip and electronic device" (环境光信号调节方法、芯片及电子设备) — priority 2019-10-25; published 2019-12-24.
§ 102 assessment: NOT prior art. This is the Chinese counterpart of the '940 patent itself (identical 2019-10-25 priority; the granted CN parent is CN 110604559 B). It appears in the citation list as a self/family citation and must be excluded from every § 102 and § 103 combination. Its presence on the face of the U.S. patent reflects family cross-citation, not examiner reliance as art.

---

C. Non-patent citation of note (not a "patent citation," but the strongest single reference in the record)

Jongpal Kim, Takhyung Lee, Jihoon Kim, Hyoungho Ko, "Ambient Light Cancellation in Photoplethysmogram Application Using Alternating Sampling and Charge Redistribution Technique," IEEE EMBC 2015, pp. 6441–6444 — published 2015 (verified).
Description: PPG readout chip with a transimpedance amplifier, LED on/off sampling, an ALC block that cancels ambient light, and — critically — a range comparator comparing the input to adjustable upper/lower thresholds (REG_th_H / REG_th_L) that drives compensator logic to increase or decrease the compensation current, implemented as an 8-bit current DAC (compensating up to 30 µA); the ALC path can be disabled via a polarity controller.
§ 102 assessment: The closest single reference in the entire record. It discloses (i) a determining circuit (range comparator) that automatically compares signal intensity against thresholds, and (ii) a cancellation circuit (logic + current DAC) whose cancellation current is automatically and dynamically set in response, with an enable/disable capability. On its face this reads on nearly all of claim 1; the residual gaps are the express "whether to turn on" decision framing and the "preset range" result limitation — which is why I treat it as a § 103 primary reference (and a near-§ 102 hit), not a clean anticipation. It is § 102(a)(1) art (published 2015, by different authors).
Sources: https://pubmed.ncbi.nlm.nih.gov/26737767/ · https://pure.ewha.ac.kr/en/publications/ambient-light-cancellation-in-photoplethysmogram-application-usin/

---

D. Ranking — most relevant prior art for US 11,326,940

1. Kim et al., IEEE EMBC 2015 (NPL) — closest overall; automatic threshold comparison + automatically adjusted current-DAC ALC with enable/disable.
2. CN 109923785 A (Goodix; Li Bo) — closest patent for the cancellation architecture (SAR loop + digitally controlled current source + comparator) and mixed-signal recovery; § 102(a)(1) art not subject to any exception.
3. CN 109863703 A (Goodix; Liu Chang — same inventor) — closest on threshold-driven automatic adjustment + subtraction, but missing the cancellation-circuit limitation and likely excepted under § 102(b)(1)(A).
4. US 9,615,427 B1 (TI) — best teaching of conditional ambient cancellation gated by thresholds, with the same noise-based motivation the '940 asserts.
5. US 2019/0298237 A1 (Nonin) — automatic, ambient-responsive control decision (but adjusts LED drive/integration time, not cancellation intensity).
6. CN 101944897 A / US 8,222,591 (Intersil) — ADA-feedback cancellation loop (SA-loop + current DAC analog).
7. US 2005/0187448 A1 (Nellcor) and US 2019/0175037 A1 (TI) — PPG/HR-domain dark-sample subtraction and dynamic adjustment.
8. Remaining references (Hamamatsu US 5,644,385; the four Fitbit citations; Philips CN 105491943 A; Google WO 2016/011173 A1; Apple CN 106559095 A; Seiko Epson JP 2019-134089 A) — background/field and corroboration art only; individually insufficient to anticipate any claim.

E. Bottom line on § 102

No cited patent reference anticipates any claim of US 11,326,940. Every independent claim requires the "ambient light cancellation circuit" whose cancellation intensity is automatically adjusted and that is conditionally turned on based on ambient intensity, and no single reference discloses the whole combination. The two references with the strongest single-reference character on the merits are the Kim et al. NPL (near-miss on claim 1) and CN 109923785 A (near-miss on the claim 7/14/15 architecture), but each is defeated — respectively — by the missing express "turn-on" decision and by the dependency of claims 7/14/15 on independent claims 1/8. The doctrine of anticipation therefore yields to § 103 across the board (as developed in the earlier obviousness section).

---

F. Caveats

1. Verified full-text/key content this session: Hamamatsu US 5,644,385; Nellcor US 2005/0187448 A1; Nonin US 2019/0298237 A1; Intersil CN 101944897 A; Google WO 2016/011173 A1; TI US 9,615,427 B1; TI US 2019/0175037 A1; CN 109863703 A (abstract); CN 109923785 A (via the earlier session's reading); and the Kim et al. NPL.
2. NOT independently re-verified this session (characterized only from the citation record / titles / abstracts): the four Fitbit citations (US 2014/0275852 A1, US 2014/0288435 A1, US 9,662,053 B2, CN 105852841 A), Philips CN 105491943 A, Apple CN 106559095 A, and Seiko Epson JP 2019-134089 A. Conclusions resting solely on these should be treated as provisional. I exhausted my search budget before fetching the Apple and Seiko Epson full texts.
3. Inventor-identity exception (flag). My conclusion that CN 109863703 A is likely excepted under § 102(b)(1)(A) rests on matching its recorded inventor (刘畅) to the '940 inventor (Chang Liu/Shenzhen Goodix). This identity should be confirmed against the Chinese application file before relying on the exception; if the inventors are legally distinct, CN 109863703 A remains § 102(a)(1) art.
4. This section refines — not contradicts — the earlier sections. The earlier obviousness section correctly noted that the § 102(b)(1) exception fails for CN 109923785 A (different inventor, Li Bo). It did not address CN 109863703 A, whose inventor appears to match the '940 inventor; that is the divergence flagged here.
5. Dates are as reported in the patent record and the sources cited; certified priority documents were not obtained.

I'll verify key reference disclosures before building the § 103 combinations, since the prior-art section flagged which references I could only read at the title/abstract level.

Obviousness Analysis — US 11,326,940 B2 under 35 U.S.C. § 103

Builds on the Prior-Art section above; that section's reference table, EPO "X"/"A" categorizations, and element-mapping are treated as the baseline and are not repeated. New here: (i) a correction to the § 102 status of one reference, (ii) verified disclosure content for the four references I previously could only read at title/abstract level, and (iii) the actual § 103 combinations with articulated motivations.

---

1. Legal framework and PHOSITA

Key structural point for the analysis: claim 1 is drafted around three functional results — (a) get an ambient signal from a PPG sensor, (b) automatically decide whether to switch the ALC circuit on based on that signal's intensity, (c) if on, automatically set the circuit's cancellation-signal intensity so the adjusted ambient signal lands in a preset range. Nothing in independent claim 1 recites a comparator, an SAR loop, a DAC, or any specific circuit topology — those are pushed to claims 7 and 14–17. That breadth is what makes the independent claims vulnerable to a two-reference combination; the dependent claims are where the circuit-specific art (CN 109923785 A, Kim) does its work.

---

2. ⚠️ Flagged contradiction: CN 109 863 703 A is probably not § 102 prior art in the US

The Prior-Art section treated CN 109863703A as the single most dangerous reference ("EPO 'X' against the entire method/chip claim set," "§ 102(a)(1) art"). I must explicitly flag that this is likely wrong as a matter of US law, and it materially changes the picture. Newly verified facts:

• Search confirms CN 109863703A's inventor is 刘畅 — i.e., Chang Liu, the sole named inventor of US 11,326,940. CN 109863703A is Goodix application CN201980000174.9, published 2019-06-07, family member WO 2020150973 A1. It is described as: "获取第一背景光信号，若第一背景光信号的强度大于第一预设阈值，则自动调节放大器增益…得到第二背景光信号；获取混合光信号…得到目标光信号…" — acquire a first background-light signal; if its intensity exceeds a first preset threshold, automatically adjust amplifier gain so the adjusted first background-light signal is ≤ the first preset threshold, yielding a second background-light signal; acquire a mixed light signal; derive the target light signal from the mixed signal and the first/second background-light signal. (Source: http://agri.nais.net.cn/patentdetails/FF87E7C0-E8B5-4733-BFD0-365A98718586.html)
• That publication date (2019-06-07) is within one year of the effective filing date (2019-10-25). Its named inventor is the same Chang Liu.

Consequence: under § 102(b)(1)(A), a disclosure made ≤1 year before the effective filing date is not prior art if "the disclosure was made by the inventor or joint inventor." On the record I can reach, CN 109863703A qualifies for that exception, which would remove it entirely as § 102(a)(1) art — and therefore as § 103 art. The EPO cited it "X" because European law has absolute novelty with no inventor grace period; the EPO's categorization therefore carries no US significance on this point.

What this does not change:

• CN 109923785 A ("Current cancellation circuit, heartbeat detection device and wearable device," pub. 2019-06-21) names a different inventor, Li Bo, is not within the § 102(b)(1)(A) inventor exception, and remains § 102(a)(1) prior art. The Prior-Art section correctly reached this conclusion for CN109923785A; the error is limited to lumping CN109863703A into the same bucket.
• Caveat / verification gap: § 102(b)(1)(A) is defeated if CN 109863703A's inventive entity includes a contributor other than Chang Liu who supplied the relevant subject matter. I have only a single-inventor listing for the CN publication; the definitive check is the front page of WO 2020150973 A1 / CN109863703A. Any § 103 ground built on CN 109863703A must be verified against that front page before filing, and a petitioner should have a fallback ground that does not depend on it.

Practical takeaway: the strongest US § 103 case is built on Fitbit + TI + Kim, not on the EPO's X reference.

---

3. Verified disclosure content for the load-bearing references

These four were read only at title/abstract level previously; I have now confirmed substance.

(A) US 9,662,053 B2 / US 2016/0302706 A1 — Fitbit ("Physiological data collection"). Its claim/abstract text recites: "light detector(s) to generate a first electrical signal while the light emitter emits the first light, a second electrical signal while the light emitter emits the second light, and a third electrical signal while the light emitter is not emitting light. The device also includes ambient light cancellation circuit(s) to generate a countering signal to counter a portion of one or both of the first and second electrical signals based on the third electrical signal." The specification adds a first switching circuit configurable between at least two configurations based on control signals, first/second sampling circuits sampling a voltage signal in each configuration, and: "an ambient light cancellation circuit for generating a first current signal to at least partially counter a first component of the first electrical signal when the first switching circuit is in the first configuration. The first current signal is based on one or more second control signals and the value of the first voltage signal sampled by the second sampling circuit." (Sources: https://patents.google.com/patent/US9662053 ; https://patentimages.storage.googleapis.com/c8/53/fe/8932daf81782e2/US20160302706A1.pdf)

(B) US 9,615,427 B1 — TI (Narayanan, Ramaswamy, Polley, Sharma). Verified: an ambient condition control circuit compares a correlation factor to a low correlation threshold value and a high correlation threshold value, and based on the factor exceeding the low threshold and being less than the high threshold, cancels the ambient light component; based on the factor exceeding the high threshold, it deactivates the illumination source; and based on the factor being less than the low threshold, no cancellation is applied (SNR is preserved and the ambient component is removed downstream by a digital bandpass filter). The patent states the thresholds "may be programmable … and may be dynamically changed," and that for PPG applications the high threshold may be set higher. Critically, TI gives the express technical reason: "applying electrical ambient cancellation adds noise and worsens SNR," so cancellation should not be applied when the ambient is uncorrelated. (Sources: https://patents.google.com/patent/[US9615427B1](/patent/US9615427B1) ; https://patents.justia.com/patent/[9615427](/patent/9615427))

(C) Kim et al., EMBC 2015 (plus its fuller journal twin). Verified abstract: a 0.13-µm CMOS PPG readout chip; against large DC offset a saturation-detection-and-current-feedback method "can compensate a current of up to 30 µA"; "an automatic emitting light compensation method"; and to remove ambient light interference, "an alternating sampling and charge redistribution technique." I additionally located the journal version — Kim, Kim & Ko, "Low-Power Photoplethysmogram Acquisition Integrated Circuit with Robust Light Interference Compensation," Sensors 2016, 16(1), 46, DOI 10.3390/s16010046 (open access, MDPI; keywords: "ambient light cancellation; automatic offset compensation; automatic emitted light compensation"). This journal article is a separate § 102(a)(1) printed publication with substantially the same disclosure and more figures than the 4-page conference paper. (Sources: https://pubmed.ncbi.nlm.nih.gov/26737767/ ; https://doi.org/10.3390/s16010046)

(D) Cross-confirmation that Kim is the industry's baseline ALC teaching. A later, unrelated patent expressly incorporates Kim as the example of ALC: "ambient light cancellation (ALC) can be incorporated to remove interference from ambient light. ALC approaches detect light both when LEDs are active and inactive… An example of ALC circuitry is disclosed by Kim et al." (US 11,471,102 B2, ¶ discussing ALC — https://patents.google.com/patent/[US11471102B2](/patent/US11471102B2)). That is admissible evidence of how a PHOSITA in this art would read Kim.

---

4. Combination A — the primary attack on claims 1, 2, 3, 8, 20

Fitbit (US 9,662,053 B2 / US 2016/0302706 A1 / US 2014/0288435 A1) in view of TI (US 9,615,427 B1), and optionally Kim (EMBC 2015 / Sensors 2016).

Claim 1 chart

Claim 8 chart (chip)

Fitbit supplies the ambient light cancellation circuit (generating a first current signal to counter the ambient component); the sampling/switching circuits plus controller and TI's ambient condition control circuit together supply the claimed "determining circuit … connected to the ambient light cancellation circuit" that "determine[s], according to an intensity … automatically, whether to turn on the ambient light cancellation circuit." TI's threshold comparison is, structurally, a determining circuit (and, per claim 14, a comparator — a threshold comparison against high/low reference values is a comparator function). Claim 20 (electronic equipment) is met by the Fitbit wearable device itself.

Motivation to combine (articulated, with rational underpinning)

1. TI supplies an express, written motivation. TI criticizes the always-on approach — "applying electrical ambient cancellation adds noise and worsens SNR" — and therefore applies cancellation conditionally (only in the mid-correlation band), withholding it below the low threshold. A PHOSITA reading TI alongside Fitbit would improve Fitbit's ALC by gating it on a measured condition, which is exactly claim 1's step (b). KSR expressly credits an express teaching in the art.
2. Power / area savings in a battery-limited wearable. Kim's design is explicitly power-driven (26 µW readout channel); Fitbit's own spec frames the device around a small battery and miniaturization. Turning the ALC off when ambient is low saves power and area — a recognized design incentive. The patent's own specification concedes this ("avoids a new interference caused by the ambient light cancellation circuit being turned on all the time, but may also save power"), which is an admission that the benefit was known and predictable.
3. Avoiding saturation/compression of the target signal. TI's stated goal is to keep the interrogation signal "of high enough quality to be processed"; the '940 background makes the identical point ("if ambient light is too intense, a received light signal will exceed a tolerance range of a back-end circuit"). Same problem → same solution.
4. Same field, overlapping inventors'/assignees' problem, predictable elements. All three references are optical biosensing front-ends; the combination is a known solution (threshold-gated compensation) applied to a known circuit (an ALC current source), i.e., KSR's paradigm of a "predictable variation."
5. Reasonable expectation of success — no new physics, no unfounded assumption; the threshold is merely a design parameter (TI even says thresholds are programmable and dynamic).

Result: claims 1, 2, 3, 8, 20 are obvious over Fitbit + TI. Claim 3 (≤ threshold → not turn on) is met directly by TI's "less than the low correlation threshold → do not cancel" branch — this is the cleanest single-limitation mapping in the whole case. Claim 2 (> threshold → turn on) is met by TI's "not less than the low threshold → cancel" branch or by US 2019/0298237 A1 (Nonin) and WO 2016/011173 A1 (Google), both of which use a measured ambient-light intensity relative to a threshold to actuate the optical front end.

---

5. Combination B — dependent circuit claims (§ claims 4, 6, 7)

Fitbit + TI (as above) further in view of CN 109923785 A (Goodix, Li Bo — "Current cancellation circuit") and Kim.

• Claim 4 (determine the electrical signal intensity; dynamically adjust the cancellation intensity so the adjusted first ambient signal is ≤ a second preset threshold, obtaining a "second ambient light signal"; cancellation intensity ≤ the electrical-signal intensity): Fitbit's ALC current is derived from the sampled voltage value of the ambient channel and counters "a portion" of the electrical signal — the countering magnitude is necessarily bounded by, and set with reference to, the ambient signal level (satisfying "≤"). Kim's current-feedback loop is the express "dynamically adjust a current until the condition is satisfied" teaching ("saturation detection and current feedback … compensate a current of up to 30 µA"). CN 109923785 A (a Goodix current-cancellation circuit for a heart-rate detection device) is the closest dedicated structure. This is a two-way combination of a compensation loop (Fitbit/Kim) with a current-cancellation circuit (CN109923785A), with the motivation being simply to implement the threshold-gated compensation of Combination A with a purpose-built current source.
• Claim 6 (cancellation intensity in direct proportion to the intensity of the first ambient light signal): met by Fitbit, whose countering current "is based on … the value of the first voltage signal sampled by the second sampling circuit" — a monotonic (and, in the ordinary linear-TIA implementation, proportional) dependence. Kim's current feedback is likewise proportional to the sensed offset. Even if not strictly linear, proportionality is a predictable design choice (KSR) once the current is derived from the ambient sample.
• Claim 7 / 15 (ambient light cancellation circuit comprising a successive approximation loop connected to a digitally controlled current source): this is the weakest limitation in the patent. CN 109923785 A (current cancellation circuit) is the best of-record candidate; Kim's charge-redistribution/current-feedback architecture is a successive-approximation-like iterative convergence, and the Intersil DAC-based offset-compensation art in this space (e.g., the digital-domain offset compensation using a stored digital code and a DAC described in Intersil's optical-proximity-sensor family — flagged: I located US 8,848,202 / CN 103105612 B as analogous, not-of-record art for this point, and the assignee/citation lineage needs verification before filing). A petitioner should expect to prove claim 7/15 with a dedicated SAR-current-source reference rather than relying on CN109923785A alone.

---

6. Combination C — dependent claim 5 (mixed-signal subtraction)

Fitbit and/or TI in view of US 2005/0187448 A1 (Nellcor) or CN 109863703A (if the latter survives the § 102(b)(1)(A) issue).

Claim 5's "mixed light signal = target light signal + first or second ambient light signal; determine the target by subtracting" is squarely met:

• TI: combined optical signal = interrogation component + ambient component; "the canceling may comprise subtracting the ambient light signal from the combined optical signal."
• Nellcor US 2005/0187448 A1: measures ambient/dark levels and subtracts the average ambient level from the detected signal (EPO "A").
• Fitbit: the first/second electrical signals (LED on) and the third electrical signal (LED off) are combined to counter the ambient portion — the same subtraction concept.
• CN 109863703A: "根据混合光信号、以及第一背景光信号或第二背景光信号得到目标光信号" — the exact "first/second background light signal" + mixed-signal → target-signal language of claim 5 (subject to the § 102(b)(1)(A) caveat in §2 above).

Claim 5 is therefore the easiest dependent claim to invalidate — four independent disclosures, all reaching target-signal recovery by subtraction. Motivation: Nellcor and TI both give the reason (isolate the pulsatile/interrogation component from the ambient contaminant).

---

7. Combination D — hardware/miscellaneous dependent claims (14, 16–19)

---

8. Claim-by-claim obviousness summary

Corollary for a defendant: there is no "safe" narrow claim to design around merely by dropping the SAR loop — claims 1, 8 and 20 do not require it, and Combinations A/C reach them with two references.

---

9. Secondary considerations (Graham factor 4)

I found no evidence of objective indicia favoring the patent, and several point the other way:

• No unexpected results. The asserted benefit (ambient treated so the back-end stays in range, plus power savings from gating) is the predictable consequence of the combination — and the specification itself states the power-saving rationale (an admission against non-obviousness).
• No long-felt need / failure of others. The industry solved this repeatedly and early: Nellcor (2004/2005), Intersil (2009/2011), Fitbit (2012–2017), Philips (2014), Google (2014), Apple (2015), TI (2015/2017), Nonin (2018), Kim (2015), and Goodix's own CN 109923785 A (2019). That is a crowded, fast-moving field, not a graveyard of failed attempts.
• Independent contemporaneous development — context only. Analog Devices' US 11,867,557 B2, "Automatic ambient light cancellation method for optical front-end," sits in the "Families Citing this family" list, i.e., ADI cited the '940 as its prior art for a 2021-filed automatic-ALC application. This is post-priority and not prior art, but it is corroborative that "automatic ambient light cancellation" was the art's ordinary next step, not an inventive leap.
• No licensing program, no NPE chain, no copying evidence (consistent with the Assignment section: sole inventor → operating-company assignee, no post-issuance transfers).

---

10. Weak points of the § 103 case (what the patent owner will argue, and how to answer)

1. "No single reference discloses the combination of (i) an automatic intensity-based decision to turn the ALC on and (ii) automatically setting the cancellation-signal intensity into a preset range."
   Answer: § 103 does not require it in one reference. TI supplies (i) with an express reason, Fitbit supplies (ii) (countering current derived from the sampled ambient value) and the ALC structure itself, and both are in the same field addressing the same problem. KSR.
2. "Fitbit's ALC switching is phase-driven (LED-on vs LED-off), not intensity-driven."
   Answer: the magnitude of Fitbit's countering current is derived from the sampled ambient value (intensity-driven); only the timing is phase-driven. And TI supplies the intensity/correlation-driven decision. The claim's "automatically" is met.
3. "CN 109863703A is the only reference that uses the claimed 'first preset threshold' language — and it adjusts amplifier gain, not the cancellation circuit."
   Answer: (a) it is likely not prior art (§ 102(b)(1)(A) — see §2) and must be abandoned as the linchpin; (b) even on its own terms, adjusting gain versus adjusting a cancellation current is an immaterial distinction on these claim words; and (c) the TI/Nonin threshold teachings and the Fitbit/Kim/CN109923785A cancellation structures supply the missing halves independently of it.
4. Claim 7 / 15 (SAR loop + digitally controlled current source) is the genuine gap.
   Reality check: this is where the patent has the most residual strength. A petitioner must find a reference or combination expressly teaching an iterative/successive-approximation convergence loop driving a digitally controlled current source for ambient cancellation. CN 109923785 A is close; Kim's charge-redistribution convergence is arguably the same idea; but I would not represent claim 7 as a high-confidence invalidation without that dedicated reference. Verification gap stated plainly.
5. § 112(b) vulnerability on claim 3 (recites "the first preset threshold," a term introduced only in claim 2) and claim 4's dependency on claim 2 only (capturing only the "> threshold" branch) are not § 103 arguments — they are IPR-eligible only as part of the claims' construction, and § 112 is not a PTAB ground (§ 311(b) limits IPR to §§ 102/103). Note them for litigation, not for the petition.

---

11. Bottom line

US 11,326,940's independent claims (1, 8, 20) are, on the record assembled here, vulnerable to a straightforward two-reference § 103 combination: Fitbit's PPG front-end with an ambient-light cancellation circuit that generates a countering current from a sampled ambient value (US 9,662,053 B2 / US 2016/0302706 A1 / US 2014/0288435 A1), in view of TI's threshold-gated decision whether to apply ambient cancellation (US 9,615,427 B1) — with Kim et al. (EMBC 2015 / Sensors 2016, 16(1), 46) supplying the power-conscious, saturation-detection/current-feedback PPG implementation and an independent ALC disclosure. The motivation is not merely "same field": TI supplies a written reason to gate cancellation rather than leave it always-on (avoid SNR degradation), and the patent's own specification supplies the power-saving rationale — an admission that the benefit was known.

• Claims 1, 2, 3, 5, 8, 9, 10, 12, 13, 14, 16, 17, 20 — strong § 103 grounds.
• Claims 4, 6, 11 — good grounds (Fitbit + Kim + CN 109923785 A).
• Claims 7, 15 — the residual strength; require a dedicated SAR-versus-digitally-controlled-current-source reference. This is the only place I would not claim high confidence.
• Do not build the petition around CN 109 863 703 A. Verify its inventive entity on the WO 2020150973 A1 / CN109863703A front page; if Chang Liu is the sole/joint inventor, § 102(b)(1)(A) disqualifies it, and any ground resting on it collapses. This is my explicit correction to the Prior-Art section, and I flag it as a factual question I could not fully close from the sources I can reach.

Confidence: high on the Fitbit and TI disclosure content (primary-source text retrieved), high on Kim (abstract + journal twin), high on the CN109863703A bibliographic/inventor data; moderate on the § 102(b)(1)(A) conclusion (turns on the full inventive entity of the CN publication, which I could not inspect), and moderate on the claim 7/15 analysis (no dedicated SAR-reference located).