Handwerker-Roberto: Quality Engineering (Step 4b)

Verdict: MINOR. The central claims are covered by deep reads, the arbiter's verdicts rest on sound calculations that I re-ran, and example.py runs. No central verdict is likely to change. The defects are: one substantive argument in the inventor's paper that no reviewer addressed (the "spherical antenna cannot emit transverse waves" step, which is his main physical reason for L.M.D. emission), some Tesla claims and related work in the folder that the arbiter and primer leave out, corrections that Step 2-3 asked for but that have not reached facts.json and claims.md, one formula used outside its valid range (skin depth in tap water), and a few reader-profile slips in the primer. All of these can be fixed by updating facts.json, claims.md, arbiter.md (QE revision section) and primer.md.

Inputs read: _text/manifest.json; every digest (13 chunks, all four documents) in full; deep/ (all three files) in full; facts.json (17 entries), claims.md, sites.md, research/2026-10-08-heos.md and the saved Heos page; physicist.md, mathematician.md, contrarian.md, arbiter.md, primer.md in full; example.py, which I ran. I also extracted pp.1-10 and 23-29 of the Investigator's 2010 deck (research/files/2011-Handwerker-Tesla-and-cold-electricity-revised-English.pdf), which no earlier step had read. Tags: [read] = project text, [web] = URL, [memory] = recalled with confidence, "calc" = computed by me in the project venv.

---

1. Coverage

All four manifest documents were digested. Step 0 skipped nothing. facts.json has one entry per document and splits the 208-page bundle into its 14 items, as the prompt requires. No defect.

2. Shortcuts

• Central claims 1-4 rest on the full deep read of the L.M.D. paper (pp.5-23); claim 5 on the H2O deep read; claim 6 on the full deep read of the prophecies. No central claim, experiment or derivation rests on a digest alone. Acceptable.
• The 13 third-party items in TESLA-THEORY-2 (Tesla 1919 and his testimony, Peterson, Trinkaus, Blazevic et al., Marincic and Budimir, Pappas, Eckardt, Trinkaus electrotherapy, Rein, Lindemann, Richards, Onoprienko) are digest only. They are not Handwerker's evidence, so this is acceptable. Two of them do bear on claims 1-2 and were not used at all (section 3, items M3-M4). That is a missed-material defect, not a coverage defect.
• Blazevic et al. pp.90-91: equations garbled in the text layer and no page images read (facts.json says so). Peripheral; acceptable, but see deep-read request 4.
• Defect (minor, sampling): the Investigator's 2010 deck has 29 pages. The Investigator and the physicist read pp.11-22 only; pp.1-10 and 23-29 were not read by anyone. I extracted them. They contain two things the reviews needed (M6 below).
• Onoprienko (Russian, pp.193-208): an English digest only. Peripheral; acceptable.

3. Missed material (from the digests and source text, absent from facts/claims/reviews/arbiter/primer)

M1. The "spherical antenna" argument (TESLA-THEORY-2 p.12), the inventor's main physical reason why a Tesla coil emits L.M.D. waves. Quoted: "a spherical 'antenna' is present which cannot emit transverse waves because of its form which mathematically annuls the opposing components of the EM transverse waves, therefore only allowing the emission of longitudinal MD (Magneto-Dielectric) waves" [read: TESLA-THEORY-2 txt, page 12; digest p.12]. No reviewer, nor the arbiter or primer, mentions it (grep for "spheric" in the five review files finds nothing). It is the one place where the paper gives a physical mechanism for the strong form of claim 1, so it needs a ruling. The mainstream point the reviewers should weigh [memory, high]: there is a real kernel. A spherically symmetric source cannot radiate transverse waves, because there is no electromagnetic monopole radiation. But the correct conclusion runs the other way. By Gauss's law the field outside a spherically symmetric charge depends only on the enclosed charge, which charge conservation fixes, so such a source emits no wave of any kind. A top sphere fed through a coil is not isolated: its charge flows in and out through the secondary, and together with the coil and the ground (or its image) it forms an oscillating dipole. That dipole has the ordinary near field, radial E_r included, and a transverse far field. The step "cannot emit transverse, therefore emits longitudinal" is a non sequitur. Note that it is an instructive one for the reader.

M2. Tesla statements that facts.json records but the arbiter and primer drop [read: Teslas-prophecies p.4-5; deep read sections 1, 4, 6]: - "a radioactive body is simply a target which is continuously bombarded ... if this ... cosmic radiation could be wholly intercepted, radioactivity would cease" (Herald Tribune 1935). This is a testable claim. Mainstream [memory, high]: decay rates are unchanged in deep underground laboratories and under heavy shielding. It belongs in claim 6 with a verdict (probably Tested, not supported, which needs a web check). - Tesla's priority claim that he discovered cosmic rays in 1896 (three 1935 articles). Mainstream credit goes to Victor Hess's 1912 balloon flights [memory, high]. This is a historical claim and should be reported as Tesla's own claim, beside the mainstream record. - The 1932 "neutrons ... carrying so small a charge" wording. The neutron itself was announced by Chadwick in February 1932 [memory, high], months before the interview, so Tesla took up a brand-new term with a different meaning. A small historical note; optional. - Handwerker's own preamble claim that Tesla "issued more than 700 patents, of which ... only about 100 are accessible" [read: p.1]. This is unsourced. The usual count is about 300 patents in some 26 countries, about 112 of them US [memory, medium]. Report it as his claim, with that count beside it.

M3. Tesla's own description of his system as non-radiating (bundle pp.45-47, testimony and Peterson): "I am not producing radiation in my system; I am suppressing electromagnetic waves"; "It is not radiated; it is conserved"; and radiated energy depends on capacity over self-induction, so he chose a large self-induction [read: digests p.45-47, p.50]. Peterson adds that the Tesla structure "inhibits the launching of electromagnetic space waves" and stores reactive energy, possibly with Zenneck-type surface waves [read: digest p.49-51]. This bears directly on claim 1. Tesla's "true wireless" is, in his own words, a stored (reactive) field and earth-current system. That fits the arbiter's near-field reading and it is also the root of Handwerker's "not T.E.M." intuition. It is pertinent and in the folder, yet none of the reviews, the arbiter or the primer uses it.

M4. Closely related work in the bundle itself: Eckardt, "What are Scalar Waves?" (2012, pp.164-172) [read: digest]. It proposes concrete transmitter designs (two antennas half a wavelength apart; a bifilar Tesla coil, US 512,340) and receivers (capacitor plates half a wavelength apart; a Faraday cage that blocks the electric part). It cites Meyl's experiment (J. Scientific Exploration 15(2), 2001) as a claimed experimental set-up, and it says plainly that "no reliable published experiments" exist for the rumoured (pi/2)c speed. The Faraday-cage receiver is the same discriminating test the arbiter proposes (the grounded screen), so it should be cited as related work. Meyl 2001 is a second claimed experiment of the same family, alongside Monstein-Wesley. Eckardt's vacuum potential wave with E = B = 0 is a pure-gauge potential with no field energy in standard theory, which facts.json already notes. No review mentions Eckardt or Meyl.

M5. The recurring "pi/2". The contrarian (correction 7) and the arbiter (correction 10) call Dollard's rumoured (pi/2)c "unrelated" to Handwerker's frequency ratio "about pi/2". Physically that is right. But the bundle has a third pi/2: Blazevic et al. give Tesla's surface-wave speed as v = c/sin(phi), "mean surface velocity c*pi/2" [read: digest p.92]. Tesla himself gave the speed of his earth-wave as 471,240 km/s, which is exactly (pi/2) x 300,000 km/s (calc) [memory, medium: Tesla, "The Transmission of Electrical Energy Without Wires", Electrical World and Engineer, 5 March 1904; needs a web check]. So the pi/2 figures probably share a source in Tesla lore, even though the ratio of two coupled-resonator frequencies has nothing to do with a wave speed (the arbiter's k = 0.397 analysis stands). The reader should get that link stated, with the physics kept separate from it.

M6. The unread pages of the 2010 deck [read: research/files/2011-Handwerker-Tesla-and-cold-electricity-revised-English.pdf, pp.24-26, my extraction]: - p.26 writes E = -grad phi - dA/dt correctly. The 2011 paper's E = grad phi (p.8), which all reviewers flag, is therefore a later slip, not a misunderstanding. Fairness requires saying so where the sign error is reported (arbiter claim 2, primer section 4 table). - p.24 gives the quaternion-versus-vector argument in full: "the sum of quaternions q1 = a+bi+cj+dk and q2 = a-bi-cj-dk gives q = 2a, which is but a scalar not equal to zero; vectors v1 = ai+bj+ck and v2 = -ai-bj-ck gives v = 0 ... results are not the same and so the involved equations ...", together with "vectors have commutative property: i.j = j.i". This is the clearest statement anywhere of why he thinks the notation matters. The comparison is not like with like: q1 + q2 keeps the two scalar parts (2a), while the vectors were given no scalar part, and a vector with a scalar part attached is just a quaternion. It belongs beside claim 2's "Refuted by calculation" as the inventor's own reasoning. - p.2: he describes himself as "a consulting professional in advanced electrotechnics, thermodynamics & energetics, also technical advisor to italian Judges" (his own description). p.5: "Experiments were realised by use of low-budget equipment ... partially using scrap parts".

M7. People and leads at the Verona event and online. The saved Heos page names the panel: Giovanni Pede (head of ENEA's low-impact vehicles laboratory), Giuseppe Buja (full professor of converters, machines and drives, University of Padua), Pietro Menga (CEI-CIVES), Bruno Giordano, Marco Tornaghi, Franco Bonfante, with Giovanni Bignami (COSPAR) as moderator. It also mentions a Heos "video notizia" [read: research/pages/heos-2011.htm; research/2026-10-08-heos.md]. Mainstream engineers were in the room, and they are possible witnesses. The arbiter and primer name only Bignami. The Facebook commenter who says he has "more of his materials" (sites.md, Facebook row) is a lead to the missing documents, and neither the arbiter nor the primer records it. The site file "Dedica Gherardo Colombo.jpg" (sites.md) is also unrecorded downstream (low importance).

M8. Small experimental details in the paper not carried forward: the pocket AM/FM radio picking up a signal out to about the Fig.3 range (E5; 2.755 MHz lies above the AM broadcast band, so what it heard is unknown) [read: p.15]; and from the 2010 deck, the vacuum-tube-driven Tesla coil and the receiving plate with a series neon lamp to earth (sites.md leads 1). Peripheral; list them in primer section 6 "What else is in the record".

4. Accuracy spot checks (against the extracted text)

Numerical checks I re-ran (calc): example.py runs cleanly and reproduces every primer number. Independent checks: near-field ratios 75 at 2 m and 1200 at 0.5 m (correct); plate field 41.8 kV/m at 100 mA and 46.0 kV/m at 110 mA (correct; the physicist's 42 kV/m for 110 mA is the slip the arbiter already corrected); Newtonian proton speed after 2.15e11 V is 21.4 c, gamma = 230 (correct); Q(Sun) = 1.66e10 C, 309 V/m, force ratio 1.1e8 (correct); Pannekoek-Rosseland net solar charge 2 pi eps0 G M m_p / e = 77 C (correct, matches the arbiter); coupling k = 0.397 for 1.522 and 0.423 for pi/2 (correct); answers to the primer's Check yourself questions 1 and 5 (correct).

Accuracy defect A1 (skin depth in tap water). The physicist, arbiter (section 2 table), primer (2.9) and example.py part 2(e) give 4.3 m for sigma = 0.005 S/m at 2.755 MHz, from delta = 1/sqrt(pi f mu0 sigma). That formula holds only for a good conductor (loss tangent much greater than 1), but example.py itself prints a loss tangent of 0.41 for this case. With the exact lossy-dielectric attenuation (eps_r = 80) the 1/e amplitude length is 9.7 m at 0.005 S/m, 1.5 m at 0.05 S/m (not 1.36 m) and 0.15 m for sea water, which is unchanged (calc). The conclusion is unchanged and in fact stronger: a vessel 10-20 cm across barely attenuates the field. The numbers and the formula's range should still be corrected, and the primer should say that the simple formula needs loss tangent >> 1. The primer's phrase "0.005 S/m (pure tap water)" should read "low-conductivity tap water".

Accuracy defect A2 (deep/H2O, minor). deep/H2O-Wireless-electricity-2955b1.md calls the 931-A "a head-on photomultiplier type", while deep/TESLA-THEORY-2 calls it "a nine-dynode side-window photomultiplier". The 931-A is a side-on (side-window) nine-stage tube [memory, high]. No review repeats the "head-on" error.

Accuracy defect A3 (physicist.md, wording). Claim 1b says "The deep read's printed lambda0 = 113.64 m is itself a 4% arithmetic slip". The 113.64 m is Handwerker's printed value; the deep read found the slip. This is a misattribution only.

5. Retention check

Material that the Investigator found or the sources contain, checked for whether the arbiter and primer keep it whatever the verdict:

• Kept well: the Verona demonstration with its stake noted and its "too early to say"; Pizzuti; the Tesla Society Switzerland copy; the Japanese blog; Arbab-Satti with their "Maxwell ... need no modifications"; Ignatiev-Leus as an open item; Monstein-Wesley, the comment and reply, and Rebilas; the patent ITVA20010043A1; the ten-plus unrecovered documents; HorizonDelta and the 2016 re-upload; the 2013 blackout with a balanced explanations table and the deciding records; Lumen notices; LinkedIn; the contrarian's "strongest honest case" (effects real and reproducible; capacitive WPT is a genuine field; E4 and E6 are sensible discriminators); "ignored, not dismissed after a test"; Tesla's quotations kept beside their verdicts. The arbiter's table of competing explanations for the silence follows SOURCE HONESTY ("Absence of evidence"). Single-source evidence is described, not labelled.
• Retention defect R1: the Tesla radioactivity and 1896 cosmic-ray priority statements (M2). facts.json lists the radioactivity statement as a key claim of the prophecies, but the arbiter and primer drop it without a verdict. Nothing suggests it was dropped because it was judged negatively; it was simply not carried forward. It is pertinent, so it is a defect.
• Retention defect R2: the Verona panel of mainstream engineers, the Heos video, the Facebook holder of "more of his materials", and the 2010 deck's vacuum-tube coil and receiving-plate-with-neon demonstrations (M7, M8) are in sites.md or the research notes but not in the arbiter or primer.
• Fairness item R3: the primer and arbiter report "E = grad phi, wrong sign" without saying that his 2010 deck has the correct E = -grad phi - dA/dt (M6). This is not suppression, but adding it is the fair statement.
• No case found of pertinent material removed because a reviewer judged it negatively.

6. Process

• Source tags: present and honest throughout. Page-image readings are tagged. [memory] carries a confidence level. The deep reader digitised Fig.3 and Fig.9 by eye and says so; the primer repeats that caveat. Good.
• Web checks: the physicist (MIT 2007, Monstein-Wesley EPL, OPERA), the mathematician (Maxwell 1865 quotation, Maxwell history) and the arbiter (6 of 8 calls: EPL items, OPERA, tap-water conductivity, US 685,957, MIT, solar net charge) checked the verdict-bearing medium-confidence items. Unchecked and stated as such: hand-in-water RF safety and meter frequency response, on which no verdict depends. Acceptable. The arbiter's tap-water citation is "via search" to a LibreTexts table; acceptable.
• Verdict scale: applied consistently in the arbiter. The arbiter fixed two mislabels: the contrarian's compass "Tested, not supported", now Never properly tested, and the physicist's non-scale "Refuted by experiment", handled through scale labels. One residual muddle: claim 5 weak form is labelled "Plausible, untested" and then "also Never properly tested in the sense ...". The primer's evidence table carries only "Plausible, untested", which is the cleaner choice; the arbiter should pick one label.
• Corrections not yet applied (defect P1). The arbiter's section 7 lists 13 corrections. Several are to Step 1 files that still read the old way: claims.md claim 3 still says "as shown by a still compass, a ferrite-core probe unchanged by the core"; claims.md claim 4 still says "read a capacitance" with no note on the mH unit; claims.md claim 2 does not say that "Maxwell incomplete" is Handwerker's reading and not his source's; claims.md's source gap lacks HorizonDelta, the missing [12] and appendix, and the unread E-book and Verona deck; facts.json key claim 8 of the L.M.D. paper still reads "as expected for waves"; facts.json and deep/H2O still give "Omdurman Univ., 2009"; the deep read's gauge passage (sections 1c and 5) is not annotated. The orchestrator should apply these, together with the stale-copy grep the arbiter lists in correction 13.
• Primer vs STEP 4: all nine sections are present in order, the answers are at the end, and example.py exists and runs. Prerequisites name textbooks and chapters (Griffiths 7, 9, 10, 11; Jackson 9; Taylor 11; Hecht 8), and these chapter numbers are correct [memory, high]. Quaternions and gauge are introduced from scratch. The worked example uses real numbers and is honest about how soft the data are. No new verdicts: the primer copies the arbiter's.
• Reader profile (defect P2, minor): these acronyms are used in the primer without being expanded: T.E.M. (first at the section 4 table; "transverse" is explained but the acronym never is), RF, LED, OPERA, GPS, CDX, S.p.A. The symbol kx in section 5 is defined (k in Step 1); fine.
• example.py: runs; output matches the primer. The only fix is the skin-depth formula (A1).

7. Remediation actions

Deep reads (in priority order; the slug is from manifest.json): 1. TESLA-THEORY-2-0d985e, pp.12-13: the "spherical antenna" argument (M1). Quote it in full with its context; say whether the paper offers any calculation for "mathematically annuls the opposing components", and how it connects the sphere's emission to the Fig.6 L.M.D. picture. Sonnet. 2. TESLA-THEORY-2-0d985e, pp.24-52: Tesla's testimony and Peterson. Extract every statement on radiation versus stored or earth-current energy ("suppressing electromagnetic waves", "not radiated; it is conserved", the 90/10 split, radiation and the capacity/self-induction ratio, Zenneck surface wave) and say how each bears on claim 1's near-field reading and on Handwerker's "not T.E.M." Sonnet. 3. TESLA-THEORY-2-0d985e, pp.164-172: Eckardt. List the proposed transmitters and receivers (bifilar coil, capacitor plates at lambda/2, Faraday cage), the Meyl 2001 citation, and the Dollard (pi/2)c remark; say whether any measured data are given; compare the Faraday-cage receiver with the arbiter's grounded-screen test. Sonnet. 4. TESLA-THEORY-2-0d985e, pp.90-92 (page images): Blazevic et al. Read the garbled equations; extract the surface-wave speed v = c/sin(phi) and "mean c*pi/2"; say whether the paper ties it to Tesla's 471,240 km/s figure (M5). Sonnet. 5. TESLA-THEORY-2-0d985e, pp.93-97: Marincic and Budimir. Exactly what they say "proved wrong" in Tesla's wireless-energy theory and what was "borne out" (ELF), with the attenuation table and efficiency figures. Low priority. Sonnet.

Not a manifest document, so listed here rather than as a deep read: the orchestrator or a reviewer should read the whole 2010 deck (research/files/2011-Handwerker-Tesla-and-cold-electricity-revised-English.pdf, 29 pp, especially pp.1-10 and 23-29) and record M6.

facts.json: apply the arbiter's corrections 3, 5, 6, 7 and 8 (Sigma unit inherited from Arbab-Satti, A/m; "asserted, not shown" for the compass and ferrite; key claim 8 "as expected for a near field" or attributed as Handwerker's expectation; mH unit; ratio 1.522 and lambda for 2.638 MHz; Progress in Physics citation). Add to the prophecies entry Handwerker's "700 patents" claim and Tesla's 1896 priority claim. Add a note to the Eckardt and Blazevic entries pointing to claim 1-2 relevance (M4, M5).

claims.md: reword claim 2 (it is Handwerker's reading, his source says Maxwell is exact), claim 3 ("asserted by", with the compass invalid at 2.755 MHz) and claim 4 (mH unit). Extend the source gap per arbiter correction 9 and note the 2010 deck's unread pages.

arbiter.md, add a "QE revision" section: - A ruling on the spherical-antenna argument (M1) under claim 1. Expected outcome: the strong-form verdict is unchanged; the argument is "Refuted by calculation" as a mechanism (no monopole radiation means no wave at all from a symmetric source; a fed sphere plus coil is a dipole). - Under claim 6: add the radioactivity statement with a verdict, after a web check of decay-rate constancy underground; and record the 1896 priority claim beside Hess 1912, as Tesla's claim and the mainstream record. - Related work: Tesla's own "not radiated; conserved" statements and Peterson (M3); Eckardt's Faraday-cage receiver and Meyl 2001 (M4); the pi/2 lore link (M5), with a web check of the 471,240 km/s figure. - Fairness note: his 2010 deck has the correct E = -grad phi - dA/dt (M6); his quaternion-sum argument (M6), with why it is not like with like. - Skin-depth table: exact values 9.7 m and 1.5 m (A1). - Claim 5 weak form: one label. - People and leads: the Verona panel (Pede, Buja and others) as possible witnesses, the Heos video, the Facebook materials holder (M7).

primer.md: - Add M1 to section 4 (a new table row, "Spherical top load") and to section 7 as a lesson: no monopole radiation, so a symmetric source emits nothing, not "only longitudinal". - Add M3 and M5 briefly to section 3.4 or section 6 "What else is in the record". - Add the radioactivity and priority lines to the evidence table once the arbiter rules. - Note the correct 2010 sign beside the "E = grad phi" row. - Fix 2.9 (the formula's range; "low-conductivity tap water"; the exact attenuation lengths). - Add the Verona panel, the video and the vacuum-tube coil and receiving-plate demos to section 6 "What else is in the record". - Expand T.E.M., RF, LED, OPERA, GPS, CDX and S.p.A. on first use.

example.py: in part 2(e), compute the exact lossy-dielectric attenuation length, alpha = omega sqrt(mu eps / 2) sqrt(sqrt(1 + p^2) - 1) with p the loss tangent, beside the good-conductor value, and comment on when each applies.

deep/H2O-Wireless-electricity-2955b1.md: correct "head-on" to "side-on (side-window)" for the 931-A (orchestrator).

Corrections to earlier documents

1. physicist.md, arbiter.md section 2, primer.md 2.9 and example.py part 2(e): the skin depth "4.3 m" for 0.005 S/m (and "1.36 m" for 0.05 S/m) uses the good-conductor formula outside its range (loss tangent 0.41 and 4.1). The exact 1/e amplitude lengths are 9.7 m and 1.5 m (calc); sea water stays 0.15 m. The conclusion is unchanged.
2. deep/H2O-Wireless-electricity-2955b1.md, p.10 section: "931-A (a head-on photomultiplier type)" is wrong; the 931-A is a side-on (side-window) nine-stage tube [memory, high], as deep/TESLA-THEORY-2 says.
3. physicist.md claim 1b: "The deep read's printed lambda0 = 113.64 m" should read "Handwerker's printed lambda0 = 113.64 m".
4. contrarian.md correction 7 and arbiter.md correction 10 ("the two pi/2 numbers are unrelated"): physically correct, but incomplete. Blazevic et al. in the same bundle give Tesla's mean surface-wave speed as c*pi/2, and Tesla's 471,240 km/s equals (pi/2) x 300,000 km/s [memory, medium; to be web-checked]. The pi/2 figures probably share a source in Tesla lore; the frequency ratio itself remains a coupling-dependent number (k = 0.397).
5. arbiter.md and primer.md (claim 2, "E = grad phi"): add that Handwerker's 2010 deck (p.26) gives the correct E = -grad phi - dA/dt [read: research/files/2011-Handwerker-..., p.26], so the 2011 sign error is a slip. A fairness addition, not a change of verdict.

---

Recheck

Method: grep and read of the current claims.md, facts.json, arbiter.md, primer.md, physicist.md, example.py and deep/ against each defect above.

Remaining items are three wording and annotation slips in secondary files (two orchestrator-owned); none changes a verdict.

Final: OPEN (3 issues)

Orchestrator fix (2026-10-08)

• deep/H2O-Wireless-electricity-2955b1.md:66: "head-on photomultiplier" -> "side-window" (with note).
• physicist.md:22: "The deep read's printed lambda0" -> "Handwerker's printed lambda0" (with note).
• Gauge passage annotation (P1): the passage on Lorenz gauge is in deep/TESLA-THEORY-2-0d985e.md (sections 1c and 5), not deep/H2O, which has no such memory statement. Added a one-line [corrected 2026-10-08 after QE] note after the passage at section 1c (~line 172) and section 5 (~line 382), citing the Arbab-Satti abstract and arbiter.md section 7 item 1.

Final: PASS