QE (Step 4b): Sykes-Chris, "Guidelines to Bucking Coils" (v2.7, 26.02.16)

QE agent, 2026-10-04. I read the following in full: prompt.txt; _text/manifest.json; the one digest; the deep read; facts.json; claims.md; sites.md; all six research notes; the three reviews (I searched them and read the parts I checked); arbiter.md; primer.md. I ran example.py. I spot-checked the extracted text layer (_text/Guidelines-to-Bucking-Coils-5ecaec.txt) and the Investigator's saved Sweet lab-notes transcription (research/files/hyiq-Lab-Notes-Transcribed-Floyd-Sweet.pdf, via pdftotext).

Verdict: MINOR

The central verdicts are sound and stay as they are. Claims 1, 2, 4 and 6 rest on the passivity and reciprocity argument, which example.py reproduces with an energy-balance check. Claim 3 rests on the deep read with page images. The defects are: - stale Step 1 files; - one deep read (pp.14-19) for a 49-page document whose circuit diagrams and experiment figures (pp.10-13, 21-24, 31-33, 40-48) were seen only through the text layer; - several items from the source and from the Investigator's files that never reached the arbiter or the primer, mostly the positive or favourable ones (see Retention); - a few small accuracy and reader-profile fixes.

None of these would move a verdict on the energy balance. Some could change the wording of the Claim 4 "conditions" verdict and the Claim 5 evidence row.

---

1. Coverage

• There is one manifest document and nothing was skipped. Coverage is complete. [read: manifest.json, _textprep.log]
• The Investigator's downloads (research/files/) are not manifest documents. They are the Sweet lab-notes transcription, "Letter to Mark from Sparky", "Nothing is Something", the hyiq.org original v2.7 PDF and the forum text dumps. The reviewers used them through the Investigator's notes. They were never digested. For the lab-notes transcription this matters: see Missed material, items M5-M7.

2. Shortcuts

• Accepted. The reviewers and the arbiter quote the text layer directly across the document. For example, "Not Free Energy" [p.48] and "100% of Rotational Shaft Energy..." [p.23] are not in the digest. So no claim rests on the digest alone. The mathematician also rendered pp.1, 8, 9 and 46 [arbiter, Source gap note].
• Defect S1 (shortcut, minor). Only pp.14-19 have a page-image deep read. The pages listed below carry figures or schematics that the text layer drops ("See below", blank lines where figures sit), and verdicts touch each of them:
• pp.40-42: the diode/square-wave circuits, Sweet's Gen-Two VTA schematic with "stud screw diodes" and an HP 211A generator, and Meyer's diode. Claim 4's "diode ... knee" condition rests on these. Partzman's diode-polarity objection is about a later schematic, so it cannot be compared with the 2016 one without seeing these pages.
• pp.45-48: the rotor experiment ("No or next to no Lenz's Law"), the lamp "swipe" experiment, and the alternator rig. These are the only Sykes experiments besides pp.14-16. Claims 2 and 4 cite them, and their figures were not viewed.
• pp.10-13 and 21-24: Sweet's 1:5 schematic, the transformer and Bloch-wall diagrams, and Sykes's "first device (2010-11)" and "second device" figures.
• pp.31-33: the loose-coupling versus tight-coupling rebuild (see M2).
• pp.25-30: the device survey figures behind Claim 5.
• Defect S2 (stale record). facts.json how_read and claims.md both still say "No deep read was on file". The arbiter flagged this, but it was never corrected. facts.json key_claims[7] and claims.md claim 6 keep the wording the arbiter corrected ("loss budget"). Arbiter corrections 1-9 have not been carried back into facts.json or claims.md.

3. Missed material

I read the digest myself and spot-read the text layer and the Investigator's files. The items below are missing from facts.json/claims.md, the reviews, the arbiter or the primer, as noted for each.

• M1. Step-up ratio as a design rule, and Sweet's turns ratio (p.11). Sykes writes: "Ideally, these devices work with a Step Up configuration. 1:3 is a common Ratio ... Current, as you may think, is not stepped down as a result of the Voltage being stepped up." He adds: "Floyd Sweet used a Step Up configuration of 1:5 ... FB1: 48 Turns P1: 240 turns" and "the VTA IS a Triode, by definition it Gates the Vacuum" [read: p.11, text lines 288-303].
• The digest records this. No review, the arbiter or the primer mentions it (a search for "step up", "1:3" and "1:5" finds 0 hits).
• "Current is not stepped down when voltage is stepped up" is a direct, testable departure from the ideal-transformer relation that Sykes quotes on the same page. It belongs in the Claim 1/4 analysis and in primer section 4.
• It also conflicts, unexamined, with the 1986 lab-notes transcription: there the excitation coil has 250 turns of #18, the output coil has 12 turns of #12, and the excitation coil is "90 Degrees Uncoupled" [read: research/files/hyiq-Lab-Notes-Transcribed-Floyd-Sweet.pdf, transcription p.30]. That is a 250:12 arrangement, not a step-up. The two may describe different VTA generations; nobody has asked.
• M2. Sykes's own loose-versus-tight comparison (pp.31-32). "I re-built this device and tightly coupled the Laminations, like a transformer ... I was not able to make the device work the same."
• This is an informal, self-reported controlled comparison of one of the Claim 4 conditions.
• The arbiter's Claim 4 verdict says "Conditions: Never properly tested (no controlled comparison)". It should say that one informal before/after comparison exists by Sykes, with no numbers, and keep "Never properly tested".
• M3. Input resonance (p.24). "I seem to get a better result if the Input is resonated for example, LC Tank ... a time constant in the Magnetic Fields of the Partnered Secondary Coils ... play a factor in the Operating Frequency" [read: p.24].
• No reviewer or the arbiter mentions it ("tank": 0 hits).
• It bears directly on primer 5.3, which resonates the output loop and shows that input current falls in a passive model. The primer should say that Sykes reports resonating the input. It could also add a sentence saying that a resonated input likewise changes current without creating power.
• M4. Self-oscillation in transformers (p.12). Sykes cites two quotes:
• Jeane Manning: "On rare occasions, Sweet saw this effect, called self-oscillation, occur in electric transformers";
• Bearden: "Sweet was also a transformer designer and expert, and he remarked that he had also observed specialized self-oscillation in certain transformers" [read: p.12].

These are absent everywhere after facts.json ("Manning": 0 hits in the reviews, arbiter and primer). They are Sweet's claimed observational origin of the effect. They also relate to the p.48 alternator self-excitation, which the arbiter rates Established. They belong in Claim 5 and in the primer's "Kept in the record". - M5. Sweet withheld a component (Investigator file, not reported by the Investigator). The lab-notes transcription, page "0014" (transcription p.46), reads:

"Checks still not in – 13 days late – Sparky explains that there is a little part in the Generator that he has withheld from disclosure – A $1.50 part from Radio Shack ... that is not a capacitor ... that puts a small µa current into the Power Coil as well as the Excitation Coil ... (he calls this unit a 'Vector Oscillator')"

This bears on the following: - Claim 5. The device as documented is incomplete by the inventor's own account, so a rebuild "as the notes describe" (the arbiter's "What would settle it") may lack a part. - The financial relationship ("Checks still not in"), which belongs in the interested-parties discussion. - The Absence-of-evidence item 3. This is a documented reason why the notes alone cannot reproduce the device.

It appears in no project file ("withheld" and "Vector Oscillator": 0 hits). - M6. Drive sources in the lab notes. The transcription says: "Power supply was from a Wavetek Signal Generator, with line power, producing a sine wave, vs. Tom's battery supplied square wave (20w?) oscillator" [read: transcription, p.6 area, lines 44-45]. - The arbiter discusses generator mains draw but says only that the model was not named. The notes say outright that the generator was line-powered, which supports the contrarian's metering point. - They also mention a separate battery square-wave oscillator of about 20 W ("20w?"), which nobody has discussed. - The arbiter's paragraph on mains draw should quote this line. - M7. The "1 mW" input is in the transcription, not only in Anor's analysis. Transcription p.30: "Operation: Oscillator put 143µa, 7v of power at approximately 400hz 1mw into the Excitation Coils ... 4.6A x 24.2V = 111.32 Watts – measured twice, observed by 3 persons." - The arbiter (table row "Input implied by 105,000-111,000x") and the primer (Claim 5 row: "about 1 mW implied on 4 June") treat the 1 mW as implied, or as Anor's figure. It is stated in the notes. The Investigator's note that the "143 uA figure differs from the transcription line '0.023mw'" is misleading, because both figures are in the transcription, on different pages and for different runs. - The ~43x discrepancy the arbiter raises therefore stands, but it is between two stated figures in the notes, not between a stated figure and an implied one. 7 V / 143 µA = 49 kΩ also agrees with the notes' "hundreds of thousands of ohm" remark only to an order of magnitude; that is worth one line. - The same page records "measured twice, observed by 3 persons" and a sensitivity to sine-wave quality and harmonic distortion. These are favourable details that should be quoted. - M8. Walt Rosenthal, "The VTA 'likes' to always see a minimum load of 25 watts" [read: p.46, text line 1189]. It is in facts.json and claims.md (Claim 4) but was dropped from the arbiter and the primer. It is one of the Claim 4 conditions. - M9. The lamp "swipe" prediction (p.47). With a big coil, a magnet and a light load, normally the lamp lights only after a delay set by the coil's impedance. In "this technology", Sykes says, "impedance reduced negligible; human eye sees light nearly immediately". - This is the document's one qualitative, testable prediction outside pp.14-16, and only the contrarian mentions it. - It belongs in the Claim 3/4 evidence row as an untested prediction with no numbers. - M10. Names in the Claim 5 lineage dropped after facts.json. - Kromrey, Markov (bidirectional transformer), Utkin, Bill Alek (SmartPAK), Mishail1971 and Hendershot (who used a bell ringer, pp.43-44) appear in no review, the arbiter or the primer. - Hooper (W.J. Hooper, whom Bearden calls "the only physicist who ever methodically investigated EM stresses ... in the laboratory", p.19) is related work under priority 3 and is absent from everything after facts.json. - The primer's "and others" should become a one-line full list. - M11. Feynman citation. The Feynman quote is on p.20, not p.21. The Guidelines give no chapter. The "Feynman Vol. II ch.15" citation comes from Sykes's hyiq.org "A Vector Potential" page [read: research/2026-10-04-hyiq.org.md item 6]. Primer section 2 ("Potentials") should cite the right source and page. Trivial point: the Guidelines call him "Sir Richard Feynman"; Feynman was not knighted [memory, high].

4. Accuracy spot checks

5. Retention

The rule: pertinent material must not drop out of the arbiter or the primer because a reviewer judged it negatively.

What was retained well. The arbiter's "What the record shows that the verdicts do not erase" section and the primer's "Kept in the record" keep: - Sykes's good practice; - the COP 1.7 attribution; - Tinman's quote; - Itsu (with its source limits); - the "deplatformed" account; - Aharonov-Bohm; - a five-part absence-of-evidence analysis.

Single-source evidence is described rather than labelled. This is good work.

Defects: - R1. Positive replication claims dropped (defect). The Investigator recorded several favourable claims, and none reaches the arbiter or the primer. Meanwhile every negative replication (dewetw 0.82, patrick1 "isnt OU", Itsu "failed") is carried into the evidence table. That asymmetry is the kind the rule forbids. The missing items are: - Sykes's 2025 statement "Marcel [FringeIdeas] has succeeded, but now has the refinement phase" (it appears only in mathematician.md); - his claim "I have shown you how to get an output, how to Generate Energy in excess of the Input"; - the forum's own "Our Above Unity Machines" list, including "Aetherholic's COP>4.0" and "Captainloz's A-R-E"; - the 2019 thread being tagged ABOVE-UNITY.

All are interested-party characterisations and should be labelled as such. They belong in the Claim 1/4 evidence row next to the negatives, with "no data seen". - R2. Sweet VTA weight-reduction test (defect). The Investigator recorded Bearden's account of a weight-reduction test: a load box of 100 W lamps, "negative AC sine-wave power", and weight recorded per 100 W step [research/2026-10-04-rexresearch.com.md]. It is a claimed anomalous experimental result. The physicist mentions it as "not read". The arbiter and the primer omit it ("weight": 0 hits in both). It should be kept in Claim 5 as Bearden's account, as an interested party, with no measurement data seen. - R3. Related-work leads dropped (minor). These Investigator leads are absent from the arbiter and the primer: - the Wiegand-wire analogue (US3820090) and the other industrial field-cooling analogues in Anor's post; - LordVëber's US4210859A; - George I. Cohn on induction; - Kapanadze, Akula, Don Smith, Gunderson and the MEG "asymmetrical regauging" lineage that Sykes himself cites after 2016.

One line in the primer's further reading or "Kept in the record" is enough. - R4. M5 (the withheld "Vector Oscillator" part) and M8 (Rosenthal's 25 W minimum load) are retention gaps as well as missed material.

6. Process

• Source tags. These are present and honest throughout. The arbiter correctly marks search-summary-only web checks as such.
• Remaining issue: the ceramic-magnet remanence of 0.38 T is still [memory, medium] in the arbiter (Claim 5, point d) and in the primer (5.5). Point (d) is part of the arbiter's argument against reading the notes as a power measurement. It should be web-checked and retagged.
• The physicist's "I know of no independent controlled test of any of them [memory, medium]" supports the "Never properly tested" verdicts for the other devices. It was not web-checked. That is acceptable because it is a statement of absence and the verdict says "in the project's material", but it should be marked as such.
• Web verification. The arbiter made 3 of its 8 allowed web calls, on generator efficiency, Goldes and Sweet's death, and Wavetek mains draw. The reviewers' web use was not audited call by call. No verdict depends on a low-confidence [memory] fact.
• Verdict scale. Applied consistently, and the split verdicts are clearly stated.
• One wording point: Claim 3's "Not testable as stated" for the interpretation means "not tested by this experiment". The strong reading ("friction on charge ... seen as heat", p.17) is testable by steady-state heating, and the arbiter's own "What would settle it" gives that test. A clearer verdict would be "Never properly tested (strong form); this experiment cannot distinguish it from inductance".
• Primer. It follows the STEP 4 order (sections 1-9, with answers at the end) and suits the reader profile well. Gaps:
• COP is used (from the evidence table on) and never defined. Define it as coefficient of performance = P_out/P_in.
• The B-H curve, its "knee" and hysteresis are used in Claims 4 and 6 without a refresher. Add a short paragraph to section 2: a B-H curve plots flux density B against magnetising field H; the knee is where the core begins to saturate; the area of the hysteresis loop is the energy lost per cycle.
• "N87 ferrite" and "MOSFET" need a few words each where used.
• example.py. It runs with the venv (exit 0) and reproduces every primer number.

7. Remediation (one round)

Deep reads (all on slug Guidelines-to-Bucking-Coils-5ecaec, most important first)

1. pp.40-42, Sonnet. - Render the page images. Describe every circuit drawn: Sykes's prescribed circuit, Sweet's Gen-Two VTA schematic with the "stud screw diodes" and HP 211A, Meyer's diode, and Melnichenko's switching. For each, give the diode polarity relative to the input-coil collapse and where the load sits. - State exactly what "operates around the knee of the BH curve, not through hysteresis" and "input coil gets choked off" are attached to. - Is the 2016 circuit the one Partzman criticised in 2020?
2. pp.45-49, Sonnet. - From page images: the rotor/bobbin experiment (p.45), including setup, magnets, rotor speed, load and any numbers. - The "swipe" lamp experiment (p.47). - The alternator rig (p.48), including the photos, how the field coil is connected, what was measured, and whether any meter readings or video links give numbers. - The Rosenthal 25 W passage.
3. pp.10-13, Sonnet. - The step-up passage and the "Current is not stepped down" claim. - The Sweet 1:5 schematic (FB1 48 turns, P1 240 turns): which VTA generation, which coil is which, and where the drawing comes from. - The "Partnered Secondary Coils" web drawing (p.13). - The 15-year study and anonymous-notes passage. - Compare with the lab-notes transcription (250-turn excitation and 12-turn output, "90 Degrees Uncoupled").
4. pp.21-24, Sonnet. - From page images: the transformer loop A/B diagram, the Bloch-wall and partnered-coil diagrams. - Sykes's first (2010-11) and second devices: construction, links table, any numbers. - The "input resonated, LC Tank" statement and any circuit shown with it.
5. pp.31-33, Sonnet. The loose-coupling device and its tight-coupling rebuild: what was built, what "work the same" means (what was observed or measured), the gap spacers, and the Jensen UDT and Bill Alek figures.
6. pp.25-30, Sonnet. The device survey: for each device (Moray, Meyer VIC, Jensen UDT, Sweet VTA redrawn, Koenig US4806834, Markov, Frolov, Melnichenko, Mishail1971, Utkin, Bill Alek SmartPAK), what Sykes shows and claims, and what evidence of performance he cites. Include the p.29 Melnichenko "not reversed" point.
7. pp.4-9 and 35-37, Sonnet. - Page images for the A/Phi definitions and diagrams (pp.4-6), the "Quantum Automobile" and Bloch-wall lamination experiment (pp.5-7), and the emf equations (pp.8-9). - The winding-direction diagrams and the Sweet standing-wave quote (pp.35-36). - Confirm the notation and unit-error findings (emf = Bvl with "l = Area").

File changes

• facts.json
• Update how_read to "digest + deep read (pp.14-19, page images) + text layer read by reviewers", plus the new deep reads.
• Correct key_claims[7] and mainstream_conflict using arbiter corrections 1 and 3.
• Add M1 (the step-up rule and Sweet 1:5), M2, M3, M4 and M9 to summary/key_claims.
• Add a second entry for the Investigator's Sweet lab-notes transcription, or note it under sources: its key numbers, including the stated 143 µA / 7 V / 1 mW input, the withheld "Vector Oscillator" part, the line-powered Wavetek and the battery oscillator.
• claims.md
• Remove "No deep read was on file".
• Reword claim 6 (arbiter correction 1) and claim 4 (alternator: "Not Free Energy", mechanical input).
• Add the step-up / "current not stepped down" statement to claim 1 or 4.
• arbiter.md: add a "QE revision" section that:
• corrects the "1 mW implied" wording (M7) and quotes the line-power and battery-oscillator lines (M6);
• adds the withheld component (M5) to Claim 5, to "What would settle it" and to absence-of-evidence item 3, and the "Checks still not in" line to the interested-parties discussion;
• amends the Claim 4 conditions verdict to mention Sykes's informal loose/tight comparison (M2);
• adds the self-oscillation quotes (M4), Rosenthal's 25 W (M8), the swipe prediction (M9) and the step-up claim (M1);
• adds the positive claims (R1) and the weight-reduction account (R2), labelled as interested-party claims;
• web-checks the ceramic-magnet remanence;
• optionally rewords the Claim 3 strong-form verdict as in section 6;
• adds anything changed by deep reads 1-7. No change to a headline verdict is expected.
• primer.md
• Define COP, the B-H curve, the knee and hysteresis loss, N87 ferrite and MOSFET.
• Fix the Feynman citation (p.20; ch.15 via hyiq.org).
• Add to section 4 a row on "Current is not stepped down when voltage is stepped up" against the ideal transformer (I₂ = I₁N₁/N₂).
• Add a note to 5.3 that Sykes resonates the input (p.24).
• Evidence table: add R1's positive claims beside the negatives, the weight-reduction account (Claim 5), the 25 W minimum load and the swipe prediction; correct "about 1 mW implied" to "stated".
• "Kept in the record": add M4, M5, M6, M10 (the full device list, Hooper) and R3.

Recheck

QE recheck, 2026-10-04. I re-read arbiter.md, primer.md, facts.json, claims.md and the deep file as they stand now (by grep and targeted reads of the passages concerned).

One issue remains (the pp.45-49 page-image deep read). It does not touch the energy-balance verdicts, and the text-layer quotes from those pages were checked in the first pass, but the gap is the one I rated first-priority for Claim 2 and 4 evidence rows.

Final: OPEN (1 issues)