Stiffler: Science Contrarian review (Step 2)

Reviewer role: Science Contrarian. Standard: experimental evidence, not consensus or reputation. Tags: [read] = the project documents or saved pages (page or figure given), [web] = Internet (URL), [memory] = recalled (confidence), [calc] = my own arithmetic from numbers in the sources.

0. Source gap and how to read this review

Every document in the folder is Stiffler's own (saved web pages and white papers, 2007-2009). Stiffler is an interested party: he promoted and sold SEC ("Spatial Energy Coherence") boards and courses under "Stiffler Scientific" [read: ECAT page; web: https://www.zpenergy.com/modules.php?name=News&file=article&sid=3221 relaying "All licenses are available from Stiffler Scientific"]. No document in the folder contains an independent measurement or replication. Figures on the ECAT page (scope test points, circuit) and SEC Special Coil (Figs 11, 13, 16) were not rendered by Step 1, so those numbers are unknown. Where a conclusion rests on [memory] alone, I say so.

The Investigator (Step I) added the later record. I keep the pertinent parts in the claim sections below: - Stiffler's own words on what CEC means (Energetic Forum thread 4067, 2009-03-12): "I do not believe in OU ... CEC ... A conversion of one energy form to another with a 'mathematical' gain. No creation of energy" [web, read by me in research/pages/ef_4067.txt; https://www.energeticforum.com/forum/energetic-forum-discussion/renewable-energy/4067-high-voltage-from-thin-air]. - His history (post #369, 2009-05-01, I read the saved page): "SEC or 'Spatial Energy Conversion' was reborn from ECAT ... The very first ECAT tries date back to 1990 and even when it looked promising the math could not support the (conversion) concept. So to avoid explaining something that seemed not to be explainable at the time, (Conversion) was replaced with (Coherence)" [web: .../page25]. - "Urban Myth" post (#2122, 2011-05-30): a CEC above 1 does not let you loop output to input, since that "would require that we were manufacturing energy, which of course is not possible" [web: .../page142, read by me]. - Replicators: Lidmotor's videos (one-wire at 10 ft and 25 ft, 2009; electrolysis 2010), forum builders (Raui, Inquorate, zeos71), brightness and voltage only, no energy balance [web, Investigator notes; I did not watch the videos]. - Death: a Ronald Stiffler, 75, Kingwood TX, died 2018-11-12 (obituary notice), a newsletter of 2019-01-24 says Dr. Stiffler "passed away last November" [web, Investigator; identity of the obituary subject is an inference from name, place and date; full obituary not read]. - Patent: "Patent Pending 61254675" (Panacea course, Scribd); no published application found [web, Investigator].

Web checks I attempted for this review (budget 6; two used): (1) a search for critiques of the Ainslie heater measurements returned nothing usable (only a forum remark that a 2002 Ainslie circuit was claimed at COP>17, unsourced); (2) a search for measured efficiency of the Strebkov single-wire line found the Strebkov-Avramenko-Nekrasov line of work and a 2019 IEEE paper on single-wire efficiency measurement (Olkhovskiy, Muellerova, Martinek) but no numbers [web: https://dspace.zcu.cz/items/ecbbaecc-629a-4d7d-8f19-0a21b92e631e/full]. So any statement below about Ainslie's own results or Strebkov's efficiency is unverified and marked [memory, low] or left out.

1. Corrections to Step 1 (and the Investigator)

1. ECAT accounting (claims.md claim 1, facts.json ECAT entry, deep read section 3). Step 1 says "the arithmetic does not close" and that, if the transformer wastes 6 W of an 8.25 W input, only 2.25 W could reach the lamp. That reading takes one ambiguous sentence at face value. The page says the 60 Hz matching transformer "wastes some 6 watts of the total input power of 8.25 watts", but its accounting is: input 8.25 W; lamp receives 12 W; transformer loss about 6 W; "total power produced by this ECAT is 18.11 watts" [read: ECAT page, lines 129-140 of the extracted text]. Read as a whole, 18 W (lamp plus transformer loss) is the power the circuit delivers into the matching transformer, and it is compared with the 8.25 W DC input. That is a coherent accounting (not double counting). The sentence "wastes ... of the total input power" is a wording slip, not a conservation violation by itself. What stays wrong or unknown: 12+6 is 18, not 18.11 (0.11 W unexplained); no measurement method is given; the 12 W is the lamp's nameplate rating, not a measurement; and 8.25 W at 12 V is 0.69 A while the page quotes "quiescent current is only 90ma" (1.08 W) [calc]. If the true input were the 1.08 W, the claimed gain would be 16.7, not 2.19; the page does not say which basis applies. The defect is lack of a stated measurement, not an internal impossibility. I withdraw "logic problem" from Step 1.
2. ECAT dates (Investigator correction accepted). I verified post #369: first ECAT tries "date back to 1990", notebooks to 1985. claims.md and facts.json should add these dates beside 1993-96. I also record two things the Investigator's note only half uses: Stiffler wrote SEC as "Spatial Energy Conversion" in 2009, and said the word was changed to "Coherence" "to avoid explaining something that seemed not to be explainable at the time". That is the inventor's own account of why the name changed; it bears on whether "Coherence" is a measurement-based concept.
3. Claim 3 (SEC14-1G2 doubling). Step 1 correctly finds the exponents misprinted (1.8225E-2 J should be 0.18225 J; 3.6450E-2 J should be 0.3645 J) [calc: 0.5 x 20e-6 x 135^2 = 0.18225; 0.5 x 10e-6 x 270^2 = 0.3645]. Add: the two cases carry the same charge (20 uF x 135 V = 10 uF x 270 V = 2.7 mC). Energy doubles because voltage doubles at equal charge, which is exactly what an ordinary voltage doubler does; an ideal doubler draws twice the charge from its source to deliver that charge at twice the voltage. The example therefore shows nothing about input versus output, as Step 1 says, and also nothing about "stacking".
4. Claim 5 (ExHeatExp). Step 1 is right about the 1.319 number. I add what Step 1 omitted from the sensitivity (section 5 below): the claimed excess equals a 3.65 K temperature over-read or a 24% input under-read, and the ratio scales directly with the mass entered (20 g gives 1.055). I also note that omissions on the heat side (glass, coil, losses) work against the claim, so they cannot explain the excess; the explanations must be on the input side or the temperature sensor.
5. Claim 6 (measurement rule). claims.md says the rule "excludes standard grounded instruments by construction". That is too strong. A grounded scope probe adds capacitance (typically tens of pF plus cable) to a node whose resonance is set by 400 pF and 23 pF inter-winding capacitance [read: Single-Wire white paper], so detuning by a grounded instrument is an expected, conventional effect [calc, below]. And the SEC15-3 exciter is powered from a DC supply (15-20 V) [read: SECExciter.htm], so its input can be measured with battery supplies and DC meters, and its output by calorimetry, with no RF probe at all. Stiffler began such a test (thermal bottle) but the page states no result. The rule does not make the claim untestable; it makes RF probing unreliable.
6. Investigator's sites.md. The statement that the Ainslie group paper "reports COP above 4.0" is from an abstract snippet on Scribd; Stiffler's name in it is unconfirmed. I keep it as a lead and do not treat it as support for claim 5.

2. Claim-by-claim review

Verdict scale: Established; Plausible, untested; Never properly tested; Tested, not supported; Refuted by calculation; Internally inconsistent; Not testable as stated; Historically superseded but instructive.

Claim 1: ECAT air-core MOSFET circuit gives power gain (8.25 W in, 18.11 W out, gain 2.194), by "Poly-Phasing" capture of external energy

a. Mathematics. No equations for "cross phasing" or the "virtual energy pump". The numbers give a gain of 18.11/8.25 = 2.195 (page 2.194) [calc]; 12 + 6 is 18, not 18.11. The accounting is coherent only as "power into the transformer divided by DC input" (see Correction 1). Not internally inconsistent; incomplete.

b. Experimental consistency. A pulsed MOSFET circuit at 4.3 kHz with ringing and a saturating core delivering a high-voltage pulse to a lamp is a classic case where input and output power are hard to measure: the DC input is filtered by a 2000 uF capacitor [read], so the supply meter sees an average, but the output is a high-frequency pulse train into a transformer and a filament. The lamp rating of 12 W holds for 120 V rms at 60 Hz; a filament's light and heat depend on the true rms heating power of the pulse train, which no instrument on the page measures. Nothing here contradicts an established result. The known failure modes are mis-measured output (rated lamp power, scope math on ringing) or mis-measured input (a DC meter on a pulsed rail whose low current and high voltage swing are misread). Which, if any, applies is unknown because the method is not given.

c. Testing. One operating point, by the inventor, no calorimetry, no named witness, no third party. Stiffler writes that Internet attempts to copy ECAT "led to total failure" because builders "used modified designs" [read]. That is his report, no attempts are named, and I found none; it cannot be counted as a replication in either direction. The "Advanced Technology Group" is unidentified; the Investigator's one query found nothing. The page reports one useful falsification-style observation: replacing the air-core coils with ferrite or metal-core transformers removed the gain "until the following design was developed" (the saturating Theta core) [read]. That is a design change the inventor says restored it; it means the claim cannot be read as "air core only".

d. If true. A source of energy other than the 12 V supply feeding a coil at kHz frequencies. The page names it "external (virtual) energy" in "the space between the power coil and the load coil". Ambient electromagnetic fields are far too weak: a 50 kW AM station gives about P/(4 pi r^2) = 50,000/(1.26e7) = 4 mW/m^2 at 1 km [calc, assuming isotropic radiation], so extra power of several watts would need about 1,000 m^2 of capture area, and at 4.3 kHz the ambient VLF level is smaller still [memory, medium]. So the unspecified source is not the ambient EM field as mainstream physics knows it. If real, it needs a new reservoir; no property of it (location, field strength, depletion) is given.

Case for the claim. It is a simple, bench-scale, falsifiable claim (a COP of about 2 at 8 W) that can be settled by a calorimetric test of the whole lamp load. The inventor says the effect depends on a specific design and was lost with conventional cores, which gives a controlled comparison he could have documented.

Against. The inventor's own admissions: early work was "empirical ... little substantiated science" [read]; his 2009 post says the "math could not support the (conversion) concept" [web]. No measurement method; the acid-test passage explains away the self-running test as "a diversion ... to hide or otherwise discredit valid research" [read], with no named person, document or event as evidence of suppression.

Verdict: Never properly tested. It was not tested by anyone else and not dismissed after a test; the inventor's own single run has no stated method. Split: the sub-claim that "external energy capture" occurs is Not testable as stated until a source is named; the measurable sub-claim (COP of about 2 at the terminals) is Never properly tested.

Claim 2: SEC single-wire coil with Avramenko plug "cohers" energy from an "Energy Lattice"; output measured as CEC = Uout/Uin

a. Mathematics. CEC% = (Uout/Uin) x 100 with U "in Joule Seconds" in the equation and in energy units in the note [read: Single-Wire white paper]; that is a unit inconsistency (joule-seconds is action, not energy). Never evaluated in that paper [read]. 130 V and 168 V at the free end from a 4.32 V p-p drive is a step-up of about 30 to 80 times [calc: 130/4.32 and 168/4.32, treating p-p loosely]; a resonant coil with a quality factor Q of a few tens does that. The LED string estimate of 168/2.7 = 62 LEDs treats a high-impedance node as a voltage source: a resonant tank can hold that voltage only while current is tiny, so the lit-LED count says nothing about power. Also "Energy Lattice" and "Spatial Lattice" are never defined [read].

b. Experimental consistency. Resonant voltage step-up, a lit LED or neon through a high-voltage point, and diode rectification of RF (the Avramenko plug is a pair of diodes) are all known behaviour. The Investigator found the same plug used in Russian single-wire patents (Strebkov, Trubnikov), by Egely, and by RSC Energia [web, sites.md]; these concern power transmission, not energy gain. What would be anomalous, an output energy above input energy, is not in any document: the white paper's own conflict check finds no result that conflicts with energy conservation. Stiffler's 2009 statement is that CEC is a conversion ratio in an open system, not creation [web].

c. Testing. Not tested in the folder. Outside it: Lidmotor and forum builders reproduced lit LEDs, high voltage, electrolysis, and one-wire transmission, with no energy balance [web, Investigator]. A replicator (Inquorate, 2019-12-01) reports that at a certain frequency "voltage and current will be out of phase. Meaning that the power calculation is always negative" [web, read]. This is an important clue to the measurement problem: at 13.56 MHz the period is 73.7 ns, and a 1 ns timing skew between a voltage probe and a current shunt is a phase error of 360 x 13.56e6 x 1e-9 = 4.9 degrees [calc]. If the real phase angle is near 87 degrees (power factor cos 87 = 0.052), a skew of 5 degrees moves the computed real power to cos 82 = 0.139 (almost 3 times too large) or cos 92 = -0.035 (negative, with the wrong sign). So both an apparent gain and an apparent negative power can come from timing errors at RF alone. The claim that SEC output exceeds input, if ever made from V x I at RF, must therefore pass a skew check. Stiffler's late figure, a CREE LED unit drawing 35 mA at 18 V = 0.63 W [web, via replicator, 2018], gives an input but no light output.

The 2010 course material claims that an SEC exciter is a "Negative Resistance" in operation [web, Scribd snippet]. Any oscillator with an active device is a negative resistance in the small-signal sense [memory, high]; this alone is not an energy claim.

d. If true. A physical medium ("Energy Lattice") coupled capacitively to a coil and diode network that gives net energy to the load. It would be an energy source with a finite size and a name; the inventor himself says that it must not be a bottomless source ("you can not just suck energy from the environment and not pass something back ... otherwise your environmental source would be ... a battery that could be drained" [web]). That position, if kept, would make the Lattice testable by drain or depletion; none was measured.

Case for the claim. Inventor honesty on one point: he rejected over-unity language and the self-running test. He built a reproducible kit and sold it (the SEC 15-20 board manual lists parts, tuning rules and limits [read]), and strangers reproduced the visible effects, which is the sort of reproducibility early claims lack. Hand capacity detunes it [read], so the behaviour is real and measurable.

Against. CEC is defined and never evaluated in the white paper. The Lattice has no properties. His 2009 admission about the failed math, and renaming "to avoid explaining" it, means the name was chosen without a theory behind it. The late-period replicators' own phase remark points to measurement artefact as a live explanation.

Verdict: Never properly tested for the claim that output energy exceeds input energy. The Lattice itself: Not testable as stated (no definable property). The observed effects (high voltage, lit LEDs, resonance step-up): Established as ordinary resonance behaviour [memory, high].

Claim 3: stacking AV Plugs multiplies "excess energy" (two double it, three triple it)

a. Mathematics. The only worked example is energy bookkeeping for two capacitors: 0.18225 J at 135 V with 20 uF; 0.3645 J at 270 V with 10 uF; the page's exponents are wrong by a factor of 10 [calc, confirmed]. Same charge, double voltage (Correction 3). The "empirical formula" for the stacking limit is not on the page. No input energy is compared.

b. Experiment. Voltage multipliers (stacked diode-capacitor stages) are standard; their output voltage is a multiple of the drive amplitude and their current capability falls accordingly. Nothing in the example conflicts with experiment; nothing demonstrates excess either.

c. Testing. None. The missing formula, and a stage count where the limit "is the minimum energy needed to excite the Lattice" [read], can't be checked.

d. If true. Each stacked stage would need its own supply from the Lattice; output proportional to stages without added input.

Verdict: Not testable as stated (the claim is expressed as "excess energy" with no input reference). The arithmetic example carries no information about excess. If the "empirical formula" appears, it becomes Never properly tested.

Claim 4: energy can be transmitted through one 12 ft wire with no return, 24 V transmitter to 48 LEDs

a. Mathematics. None given; no frequency, inductance, current, power. The coil is 143 turns tapped at 37 on a 12.5 mm Lucite rod, with 'top hats' made of stainless condiment cups [read].

b. Experiment. Single-wire energy transfer is not forbidden. A top hat is a capacitor plate; displacement current flows through the capacitance to the surroundings (ground, room, the other top hat) and closes the loop, so "no return conductor" means "no metallic return conductor" [memory, high]. The guide's own observations (hand capacity detunes it; metal near the wire reduces efficiency [read]) fit that picture. Tesla used this approach in the 1890s and the Strebkov group published and patented kilowatt-scale single-wire resonant lines [web, Investigator; efficiency numbers not found by me].

c. Testing. The Lidmotor builds at 10 ft and 25 ft around a wall are independent visual replications of the effect [web, Investigator; I did not watch them]. No power, efficiency or loss figure exists in the folder; the guide defers measurement to "Measurements for SEC Exciters". The demonstration video y9aJohml130 is removed [web].

d. If true. Nothing changes: it's capacitive coupling plus resonance. What would be new is a high efficiency over metres with an "unusual" field strength; the guide's warnings of headache, tingling and electronic damage within the field [read] are plausible for strong RF near fields and should be treated as a safety, not a physics, statement.

Verdict: Established for the weak form (lighting LEDs through one wire with a capacitive return), as ordinary resonant capacitive transfer. Never properly tested for the strong form (any efficiency or gain figure, or transmission beyond what a displacement-current picture predicts); no one has published the power in and out.

Claim 5: in Stiffler's calorimetry of the Ainslie MOSFET heater, heat in the water exceeded measured electrical input (CEC 1.319), and returned energy is insignificant

a. Mathematics. The one result is internally consistent: Q = c m dT = 4.187 x 25 x (41.25 - 26.15) = 1580.6 J; Sum(V I t) = 1198.759 J; ratio 1.3185 [calc, matches the Fig 15 screenshot; read: deep read]. Mean input 2.0 W, heating 2.63 W [calc]. The input is accumulated once per second from voltage and current readings ("Read Interval 1 s") [read: Fig 15], while the footnote says energy readings with AC components "are taken by signal integration at a 40 MHz sampling rate" [read]; the page does not say which was used for Fig 15.

b. Consistency with experiments. What would be needed to turn 1.319 into 1.0 [calc]: - Temperature: the rise would have to be 11.45 K instead of 15.10 K, an over-read of 3.65 K. A temperature sensor LM335 gives 10 mV per K, so 3.65 K is a 36.5 mV error. Rectified RF pickup in a sensor lead near 1.49 MHz ringing and a 149 V spike could easily be tens of millivolts; the toroid filter in Fig 18 suggests Stiffler suspected this [read: Fig 18 image; purpose is inference]. The thermometer sat in the top ~5 mL of an unstirred tube heated from the outside wall by the coil; hot water rises, so the top reads above the mean [read: no stirrer]. Either can inflate dT. - Mass: with 20 g (the page says 20-25 mL) the ratio is 1.055 [calc]. - Input: an under-read of 24% (1580.6/1.0 vs 1198.8) would do it. Supply voltage sags to 16.97 V at times [read: Fig 15], so a reading of "20.00 V" and current sampled once per second of a pulsed load can plainly miss structure. Items that cut the other way (so they cannot explain the excess): heat in the glass, coil and diode (epoxied to the tube [read: Fig 17]), and losses are not counted in Q, making true heat higher. So the result is not dismissible by "he forgot losses"; it has to be explained by sensor, mass or input measurement, none excluded.

c. Testing. One run, one screenshot; no repeat, blank (the same energy delivered as plain DC to the same coil), calibration constant, or error bars. With cups, CEC>1 could not be sustained above 42 C and "gradually declined to where it leveled at around 83%" [read]; Stiffler attributes this to heat loss (consistent with a poor enclosure). The page ends unfinished, with a "second inductor" added and no result [read]. The Ainslie group's paper reports COP above 4.0 [web, Scribd abstract; the group originated the circuit and is an interested party; Stiffler's role unconfirmed]. I cannot confirm any replication of Ainslie's own claims with a source; I leave it blank.

He argues that returned energy to the battery is insignificant because the freewheeling diode D1 across the coil keeps the collapsing current in the coil loop, and "I have confirmed" it heats the coil and diode, not the battery [read]; he shows no waveform or number for it. In his own assumptions the pulse is under 1 kV, so the MOSFET body diode does not conduct; the measured spikes are about 58 V and 149 V [read: Figs 13-14]. This argument is conventional circuit theory and agrees with a standard analysis: with a diode across an inductor, flyback energy circulates and is dissipated in the loop [memory, high].

d. If true. A sustained heat output above measured electrical input would need either an input measurement error or a new energy source. Stiffler does not claim a new source; he writes of neither "COP" nor "over unity" on this page [read].

Case for the claim. Better than anything else in the folder: a stated apparatus, a calibration statement (LM335 "calibrated before each run to a NIST traceable standard"), an energy-integration method with 40 MHz sampling and low-inductance shunts, and a test he reports honestly including the failure in cups. This is the right kind of experiment (calorimetry on a pulsed circuit), and it is a test anyone can rerun in a day.

Against. Single unrepeated run; the integration method for the reported figure is not identified; the sensor was in a strong RF field with no stated check; the page has no date and is unfinished; the 40 MHz sampling of a 1.49 MHz ring is about 27 samples per cycle, fine, but a 600 s total at 1 s intervals is a different method.

Verdict: Never properly tested. A first, unreplicated calorimetric run that gives a ratio above 1 and was not followed by a documented control; the possible artefacts (sensor RF pickup, thermal gradient, mass, input integration) were not excluded. Not "tested and failed" and not "dismissed without testing". Returned-energy sub-claim: Plausible, untested (standard circuit theory, but no waveform shown).

Claim 6: SEC exciters can be measured only with battery-powered, ungrounded instruments behind RF filters; grounding "will detune the AV Plug to where it no longer will interface with the Spatial Lattice"

a. Mathematics. Stiffler's equations are mostly standard: P = (Vr/R)^2 R for a load, LED power = V_LED x n x Is plus 0.6 V diode drop x Is, energy by integrating V x I across a 1 ohm shunt [read: Measurement Methods]. The claim "UCEC > 1 implies PCEC > 1" is true only if the two measurement intervals are equal (Step 1 notes this). Eq 9 is dimensionally wrong as printed (Step 1). The SEC Special Coil impedance tables use sqrt(|X| + R) instead of sqrt(R^2 + X^2) [calc, deep read]; the minima fall in the right places but the Z values cannot be used in power or Q calculations.

b. Experiment. The rule is physically meaningful: a node resonant with 400 pF in parallel with a 10 pF probe is shifted by about 2.5% in capacitance and about 1.2% in frequency [calc], and the high-Q response is sharp ("one sharp spot" in the one-wire guide [read]), so a grounded instrument detunes a high-impedance RF node. His warnings that field pickup can give "meaningless relative readings" [read] are what any RF engineer would say about probing a strong near field. The mechanism as stated ("Spatial Lattice") is an interpretation of an ordinary loading effect.

c. Testing. The rule itself was never tested as a claim; it is a lab practice. The tests in the folder never use a calorimeter on the output, apart from the thermal-bottle comparison against a resistor at equal input power, described but with no result on the page, and a LED/black-plate test with no temperature rise [read]. A no-result result is itself a result to flag: the one test that could have shown excess heat produced none that is reported.

d. If true. If the Lattice interface is real and grounding breaks it, then the claim is unfalsifiable by any grounded instrument, but a thermal bottle with battery supply does not require grounded RF probes and is a legitimate test the inventor himself proposed.

Verdict: Plausible, untested as a statement about instrument loading (conventional). The Lattice reading of it: Not testable as stated. If the rule is used to excuse the absence of a measured energy balance, then the claim for gain remains Never properly tested.

Claim 7: Stiffler's reports on testing: copies of ECAT failed because builders changed the design; Naudin's TEP fell short; the "acid test" is a diversion by suppressors; the LED/black-plate and thermal-bottle tests

a. Mathematics. None. The door-and-wind analogy [read] is a valid open-system point: a system fed by an external source (a solar cell, a wind turbine) can show COP above 1 relative to its electrical input without violating anything; the analogy then claims the generator "won't work" if it is used to keep the door open, which is the self-running version he rejects. The 2011 post says the same: CEC>1 does not allow looping [web].

b. Experiment. The argument is correct as thermodynamics: open systems need not be closed-loop self-running. But the claim then requires that the "external" energy is measured at its source; a heat pump has COP 3-4 with the source known (outside air) and no one calls it a perpetual device. For ECAT the source is not identified, so the "open system" defence has no content until it is.

c. Testing and rejection. Nothing was rejected by an institution. No journal, laboratory, agency or critic rejected Stiffler's results on the record I have: he did not submit papers, the planned SEC paper (ZPEnergy 2010) was never found, and no patent was published. The "suppression" statements name no person or document [read]. The Investigator found Google removal notices (Lumen 75850580 and 95550170) that cannot be read; who sent them is unknown, and I draw no conclusion [web]. So this is a case of never tested by others and not dismissed after a test, because almost nobody outside the hobbyist community looked. The most likely reason the record goes quiet is the inventor's death in November 2018, not secrecy (Investigator's analysis; competing readings: lapsed provisional patent, unpublished paper; the USPTO record for 61/254,675 would decide the patent question).

d. If true. For the suppression claim: a documented act (a withdrawn grant, a refused paper, a threat); none is on record. For the TEP comment: a statement by a party that claims priority over TEP.

Case for. The inventor is honest about not claiming a self-running device and says so repeatedly, which makes his claim narrow and testable (COP above 1 with continued input). Against. Statements about other people's failures and about suppression are unsupported assertions.

Verdict: Not testable as stated (suppression and the "Free Energy folks always changed the design" statements cannot be checked: no person or document is named). The testing reports are the inventor's own: Never properly tested.

3. What I would keep in the record whatever the physicists conclude

• The ExHeatExp run is the best-documented experiment here, and no one has repeated it; keep the Fig 15 numbers and the sensitivity table.
• Single-wire resonant transfer is real engineering with a published and patented literature (Strebkov et al.), which means claim 4 is partly mainstream-adjacent; Stiffler's version adds only the Lattice interpretation.
• The inventor's 2009 words: "I do not believe in OU ... No creation of energy", and the admission that the original conversion math "could not support" the concept. Both belong in front of the reader; one rules out a perpetual machine, the other shows the term "Coherence" was introduced to avoid a failed explanation.
• Replicators' reports (Lidmotor, Raui, Inquorate, zeos71) all report visible or voltage effects, none an energy balance, and one points to the phase problem. The absence of a single energy balance by any replicator in 9 years is itself an anomaly: the effect is easy to build and easy to measure by calorimetry, and no one on record did it.
• Absence of evidence: the record ends in mid-2018. Inventor's death in Nov 2018 is the best-supported explanation; the demo video, the websites (stifflerscientific.com, drstiffler.com) are gone; the Wayback Machine was blocked during the Investigator's run (retry).

4. What would settle it (single tests)

1. Claim 5: repeat ExHeatExp with a stirred cell, two independent sensors (one shielded and one optical or non-electronic), a known mass weighed, a DC blank at equal power, input by integrating V x I from a fast digitizer on both the pulsed run and the blank. If CEC for the blank equals the pulsed run, no effect.
2. Claim 1: put the ECAT lamp load into a calorimeter or measure the lamp light against a mains-driven reference of the same lamp; measure input with a battery and a DC integrator.
3. Claim 2: run an SEC15-3 on a battery inside a thermal bottle against a resistor of equal input power (Stiffler's own proposed test), long enough to reach steady state. Add a phase check with matched-delay probes (skew under 0.1 ns at 13.56 MHz) before accepting any RF V x I.

5. Verdict summary

1. ECAT power gain 2.194: Never properly tested (accounting coherent but unmeasured; "external energy" part Not testable as stated).
2. SEC / Energy Lattice / CEC: Never properly tested (gain); Lattice Not testable as stated; ordinary resonance effects Established.
3. Stacking AV Plugs multiplies excess energy: Not testable as stated.
4. One-wire transmission: Established (weak form, capacitive return); Never properly tested (any gain or efficiency).
5. Ainslie heater calorimetry CEC 1.319: Never properly tested (single unreplicated run; artefacts not excluded); returned-energy sub-claim Plausible, untested.
6. Measurement rule: Plausible, untested as instrument loading; Lattice reading Not testable as stated.
7. Reports on testing and suppression: Not testable as stated; no rejection by any institution on record.