Correa (PAGD / "XS NRG" energy conversion): PhD physicist review

Reviewer role: PhD physicist, mainstream view. Inputs: facts.json, claims.md (Step 1), checked against the source text.

Source gap (repeated from Step 1, and what I did about it)

• Only the rexresearch.com printout "Paul & Alexandra Correa _ Pulsed Abnormal Glow Discharges (PAGD).pdf" has a text layer. I re-extracted it with pdftotext -layout and read the key passages myself: press release, "Power From Autoelectronic Emissions" (lines ~320-390), the patent text on Tables 7/8 and the four measurement methods (~1590-1860), and Tables 12-14 plus the reaction-force worked example (~4165-4230). [read]
• The two patent PDFs (US 5,449,989; US 5,502,354) are image-only scans. I did not re-render them; I rely on Step 1's reading of page 1 and on the same patents' text as bundled in the printout. US 5,416,391 exists only inside the printout.
• Every number below is the inventors' own. No independent test report, original lab notebook, or oscillogram is in the folder. Conclusions about whether an energy surplus is real rest on the inventors' text and on physics arguments, not on any independent measurement.

Corrections to Step 1

1. facts.json describes Table 7 as the main breakeven evidence ("mean DC volts and amps ... authors admitting that battery state/time of delivery was ignored"). That is right for the third method, but the printout reports four methods (open-circuit voltage, time-integrated negligible-load discharge, operational power, and pre/post resistive discharge curves, Table 8). The inventors call the last two the most reliable. They give group 6 as 483% (37.7 Wh lost by the drive pack, "182.2 kWh" gained by the charge pack; "kWh" is clearly a typo for Wh) and say the methods differ by up to 10-fold. The claims.md range "218 to 563%" therefore mixes a single method; Table 8 is the relevant one. [read]
2. claims.md claim 2 and facts.json say ion energies of "about 0.5 MeV" were reported. The text says they were calculated, and the calculation is circular: 86.4 J (mean net reactor pulse energy) divided by 1.3e15 ejected Al atoms gives 4.8e5 eV each. It is the energy assumed to be delivered, spread over the measured mass loss, not a measured ion energy. No ion energy analyser or time-of-flight measurement is cited. [read; I reproduced the division: 86.4 J / (1.3e15 x 1.602e-19 C) = 4.15e5 V, i.e. about 4e5 eV, same order as their 4.8e5.]
3. claims.md claim 3 joins two different quantities ("218 to 563%" battery energy ratio, and "3 to over 50 times" pulse energy). The latter (Table 14) comes from oscillographic readings across series capacitors (5.3 uF or 17.5 mF), a different method. Both are the inventors' own.
4. Missing from Step 1: the printout's own statement that the open-circuit-voltage method "overestimates net power ... by up to 10 fold" and that the batteries delivered only 50-60% of rated capacity. These matter for interpreting claim 3.
5. Minor: facts.json lists the "pulse energy" example as "mean net pulse output 86.4 J, ... 245.2 dynes (reaction balance)". Correct as written; I only add that the force (143.6 dyn) is momentum divided by the 40-50 ms pulse time and the factor 1.39 (reproduced: 5.8e-8 g x 1.7e8 cm/s = 9.86 g cm/s; 9.86 / (1.39 x ~0.049 s) = ~145 dyn). [read + my arithmetic]

Claim 1: self-pulsing cold-cathode tube; emission at far lower field than Fowler-Nordheim (FN) predicts

Fowler-Nordheim: the quantum-tunnelling formula for electron emission from a metal at a very high surface field (of order 1e9 V/m).

a. Mathematics. The patent's own argument compares the average gap field (5,000-30,000 V/m; Table 12) with the FN requirement (>2e9 V/m) and infers an enhancement factor beta of 1e6 to 1e7. [read] That comparison is a category error. FN concerns the local field at the emitting surface; the average field in a 5-8 cm gap is irrelevant unless one specifies how microscopic protrusions concentrate it. The text also says the discharge runs at 1-2 Torr of air in the worked examples (lines ~949, 977). [read] At such a pressure and gap (pd about 10 Torr cm), air breaks down by gas (Townsend/Paschen) avalanche at a few hundred volts [memory: high confidence, Paschen minimum of air is ~330 V]. The 560 V quoted is of that order. So no FN failure has been demonstrated. The claim of a "positive I/V^2 slope" (inverse FN plot) is likewise what one expects from gas multiplication, not field emission.

b. Experiments. Relaxation oscillation of a DC-fed glow discharge with a capacitor and series impedance is a textbook phenomenon (negative differential resistance in the abnormal-to-arc transition) [memory: high]. Field enhancement from microprotrusions and adsorbates, with beta of 50-500 typical, is standard [memory: high]. Nothing here conflicts with experiment except the FN-failure reading.

c. Tests. Only the Correas. No independent replication found in the folder; a web search (aetherometry.com, rexresearch, aetherenergy.com) returned only the Correas' own pages and no independent test. [web: https://www.aetherometry.com/Labofex_Plasma_Physics/cat-Lab.php, https://www.rexresearch.com/correa/correa.htm; absence of replication is "not found", not proof of absence]

d. If true. A new low-field emission regime would not break any law; it would just need a better model of microprotrusion and gas-phase triggering. Minor change.

Verdict (split): - Weak form (a DC-fed vacuum/low-pressure tube spontaneously self-pulses): Established. - Strong form (FN is violated / a new emission mechanism): Not testable as stated, because the "violation" rests on comparing mean gap field to local surface field; once beta is allowed, the claim has no discriminating prediction. Never independently tested.

Claim 2: 0.5 MeV ions and 100-300 dyn reaction forces, 100x vacuum arcs, as Aspden's Law predicts

a. Mathematics. The arithmetic is internally consistent (checked above: 5.8e-8 g, 1.3e15 Al atoms, 1.7e8 cm/s, 143.6 dyn). The problem is energy bookkeeping. 1.7e6 m/s for an Al ion is 4.1e5 eV (0.5 m v^2, m = 4.48e-26 kg), matching their figure, so v was derived from the same circular energy. A singly charged ion gains at most e x V from the electric field. The text gives reactor pulse voltage Vp of 100-500 V (line ~4210) and says Vp cannot exceed breakdown voltage. So an ion can carry about 500 eV, three orders of magnitude below 4.8e5 eV, unless it is multiply charged by a factor of ~1000, which is impossible for aluminium (Z <= 13). Check: 1.3e15 ions x 500 eV x 1.602e-19 = 0.10 J, versus the claimed 86.4 J. The authors recognise the voltage limit and call it "current amplification", which does not change the energy-per-charge limit for ion acceleration in an electric field.

b. Experiments. Tanberg (1930) measured Cu ion velocities of about 1.6e4 m/s in vacuum arcs; typical cathode-spot ions have 1-3 charge states and tens of eV [web: https://www.osti.gov/servlets/purl/810482; search summary also at https://www.osti.gov/servlets/purl/836650]. Cathode jets in vacuum arcs are well explained by plasma pressure and ion acceleration in the spot (including a "hump" potential). Correa's claimed 1.7e6 m/s is about 100x faster than any measured vacuum-arc ion jet, and no direct ion measurement is given. The reaction force F = momentum / pulse time, using mass loss estimated metallographically, is also a known route to macroscopic forces.

c. Tests. Reaction balance of 245 dyn is the inventors' own. Graneau-type claims (water arc explosions, Ampere longitudinal forces) have been disputed; standard Lorentz-force and thermal/pressure explanations were offered by critics, with Graneau maintaining otherwise [web: https://www.sciencedirect.com/science/article/abs/pii/0375960189900893 for a longitudinal-force wire-arc paper; search summary shows the debate is unresolved only in the Ampere-law community]. Aspden's Law (force depending on ion-to-electron mass ratio, ~100 i^2) is, as far as the text shows, a fit made to Tanberg-type results by Aspden himself; the printout cites no independent confirmation. [read; independent status: unknown]

d. If true. Newton's third law (as applied to the discharge plus field) and the Lorentz force law would need a non-standard term; ion acceleration by a 500 V field to 0.5 MeV would violate energy conservation unless an energy source exists.

Verdict (split): - Forces of 100-300 dyn from a pulse that ablates ~6e-8 g at ~1e6-1e8 cm/s: Plausible, untested as a momentum-bookkeeping statement; the numbers are in the right range for plasma jets but the velocity is unmeasured. - 0.5 MeV ion energy: Refuted by calculation (energy-per-charge limit above: 500 eV x 1.3e15 = 0.1 J vs 86.4 J) unless a separate energy source is shown; the quoted value is a derived, not measured, quantity. - "As predicted by Aspden's Law, not Lorentz": Never properly tested (no neutral test; Aspden's law is not in the standard literature).

Claim 3: double-ported circuit outputs more electrical energy than it consumes (218-563%)

a. Mathematics. The definition BE = delta E (charge pack) / delta E (drive pack) x 100 is simple and consistent. The inconsistency is in how the two energies were obtained. They are not metered energies but inferences from lead-acid gel-battery discharge curves, before and after a run, to arbitrary cutoffs (115 W for the 46-cell drive pack, 90 W for the 29-cell charge pack). [read] For group 6 the drive pack lost 37.7 Wh (1.8% of its 2.066 kWh full capacity) and the charge pack gained 182.2 Wh (13.8% of 1.320 kWh). Absolute numbers are small relative to a battery's capacity and well inside the capacity ambiguities they themselves report (effective capacity 50-60% of rated; methods disagree by 10x; "actual performance ... in terms of time, time of delivery and time of capture was also ignored"). [read]

b. Experiments. No conflict with a tested law other than energy conservation itself. Known battery effects that can fake a gain: lead-acid recovery after rest (diffusion, voltage relaxation), differing state-of-charge/discharge-rate histories, charge-acceptance dependence on pulsed charging, the arbitrary cutoff power, and comparing two different packs with different loads (R2 vs R3). The text shows both packs have strong hysteresis ("significant memory effect") and uses the 1-minute-on / 15-minute-rest sequence that gives the highest BE (Table 8 groups 1-2): the longest relaxation periods give the largest apparent gain, which is what a battery-recovery artefact predicts. [memory: medium on the general battery physics; the text supports the pattern.]

Real-time meter data (Fig. 20): drive pack 36.6-57.8 W, charge pack 146-400 W, from DC meters on pulses of 5-60 per second. [read] A single meter sampled a few times per second on a pulsed signal cannot give the mean of V x I unless sampling is synchronous; the 2.7x spread suggests aliasing. The authors fit "485% +/- 18%" to n = 64 samples, which is a statistics statement about meter readings, not an energy measurement.

c. Tests. Only the inventors' own. The same text proposes a self-looping circuit (Fig. 12 swaps drive and charge packs). If the gain were real, the cleanest demonstration, a closed loop that runs unattended or a continuous run with the same pack, would settle it. In the text I read I found no report of such a closed-loop run. [read; absence in the text read, not proof none exists]

d. If true. Conservation of energy would have to be extended by an external source ("the circuit clearly draws on a further source of energy input", their wording). That is a revision to thermodynamics and QED vacuum physics, not a tweak.

Verdict: Never properly tested (no neutral, controlled test; no closed-loop run, no calorimetry, no independent lab). The measurement chain leaves ordinary battery artefacts open. Not "refuted by calculation", since the claim is an experimental one and no budget calculation applies to the battery accounting.

Claim 4: surplus energy comes from vacuum / zero-point / aether energy

a. Mathematics. The text offers no equation connecting the vacuum to the measured output. It cites Aspden's "dynamic zero-point field" and Shoulders' EV work; the quantitative link is a comparison of energy densities (2.8e9 J/cm^3 in the cathode vs "three orders" below Aspden's vacuum value). [read] That is a numerical coincidence argument, not a derivation. Also "dynamic capacitance 0.2 to 80 mF" for a 2 pF tube (line ~4215) is an ad hoc quantity computed by dividing claimed energy by V^2, so it assumes the answer.

b. Experiments. Zero-point energy is the lowest state; you cannot extract work from it in a cycle (a lower state cannot be tapped). The Casimir effect and vacuum-fluctuation effects are real but are not a power source [memory: high]. No experiment supports macroscopic extraction.

c. Tests. None; the proposal is untestable until claim 3 is demonstrated.

d. If true. A violation of the second law and the definition of vacuum as ground state.

Verdict: Not testable as stated (the explanation has no measurable prediction that differs from "the measurement is wrong"). It is moot until claim 3 is established.

Claim 5: the pulsing tube as a DC-to-AC inverter driving induction motors (US 5,416,391)

a. Mathematics. A relaxation oscillator charging a capacitor through a tube, capacitively coupled to a motor, is consistent; speed set by pulse frequency follows from f = 1/(RC-type time constant). Nothing requires over-unity. Step 1's note that claims were not examined in detail stands; I also did not read the claims.

b. Experiments. Gas discharge relaxation oscillators and spark-gap inverters are old technology [memory: high]. Efficiency would be poor (high-voltage drop, current-limited), but nothing conflicts with experiment.

c. Tests. Demonstrations are by the inventors; the physics is routine and needs no special test.

d. Nothing changes.

Verdict: Plausible, untested (as a working device with the stated performance; mechanism Established). I do not give "Established" to the patent's performance (efficiency, torque) as no numbers were independently tested.

Claim 6: "Aether Motor" powered only by orgone accumulators / ground

a. Mathematics. No equations or schematic in the files. Cannot check.

b. Experiments. Reich's orgone accumulator effects (temperature anomalies, etc.) have not been reproduced under controls [memory: medium; I did not verify this by web]. A device supplying electrical power with no external input contradicts energy conservation unless the ground/atmosphere is a source (atmospheric potential gradient of ~100 V/m can supply microwatt-level power only; known).

c. Tests. Only Axelrad's testimonial; he states he is not trained in plasma physics. [read]

d. If true. New physics of orgone energy; a perpetual-motion machine otherwise.

Verdict: Not testable as stated (no technical content). If a specific device were built it would fall under claim 3's standards.

Summary table

Web checks done (budget ~6)

1. Vacuum-arc ion velocity / Tanberg 1.6e4 m/s, charge states: confirmed [web] https://www.osti.gov/servlets/purl/810482 and https://www.osti.gov/servlets/purl/836650.
2. Independent replication of Correa: none found [web] aetherometry.com / rexresearch.com pages are the Correas' own.
3. Graneau / Ampere longitudinal forces: still contested [web] https://www.sciencedirect.com/science/article/abs/pii/0375960189900893. Not web-checked, still [memory]: Paschen values for air (high confidence), lead-acid recovery effects (medium), Reich accumulator replication record (medium), Aspden's law not accepted in the standard literature (medium). Conclusions that depend on them are marked above.