JLN Labs (Naudin's own site), 2026-10-09 (Investigator, Step I, run 1)
Site: http://jlnlabs.online.fr (also http://jnaudin.free.fr, which redirects or mirrors). Fetched by script (curl) to
research/pages/ and read as text. Pages that return a 2,209-byte page are free.fr "404" pages.
Queries / paths tried: cfr/index.htm, cfr/htmls/cfr_data.htm (404), cfr/html/cfrdatas.htm, cfr/html/calib.htm, cfr/html/cfr33.htm, cfr/html/cfrtpwr.htm, mahg/index.htm, mahg/logbook.htm, mahg/mahg1.htm, mahg/tests/index.htm, mahg/tests/lrtests.htm, mahg/tests/pultests.htm, mahg/tests/mahg2e.htm, mahg/tests/maghwr.htm, meg/index.htm (404), jnaudin.free.fr/meg/meg.htm, meg/megv2.htm, meg/megjln01.htm, meg/megnot01.htm, meg/megv3.htm, megv30, megv31, megv32 (all 404).
Result: positive. Several of the data pages that claims.md lists as missing are still online.
1. CFR (claim 4): data and calibration pages
- Results index
cfr/html/cfrdatas.htm(created 2003-05-08, updated 2006-03-22) [web] http://jlnlabs.online.fr/cfr/html/cfrdatas.htm: Naudin states he replicated the Mizuno-Ohmori experiment on 7 May 2003 following Mallove's protocol (amasci.com/weird/anode.txt and Infinite Energy vol. 4 no. 20, 1998), and calls it "fully replicable". The page names his references: Mizuno et al., Jpn. J. Appl. Phys. 39 (2000) p.6055; Mizuno, Akimoto and Ohmori (2003, anomalous hydrogen); Kanarev; the Thermacore atomic-hydrogen paper; patents WO0025320 (Eccles), WO0126117 and WO9949471 (Mizuno), JP5027062. It quotes Mallove that the anode effect was reported by Kellogg (J. Electrochem. Soc., April 1950, p.133). The run table itself is an image, not text, in the saved copy. - Calibration
cfr/html/calib.htm(2003-05-08 / 05-22) [web] http://jlnlabs.online.fr/cfr/html/calib.htm: a resistive electric heater in the same beaker and 0.2 M K2CO3: input 297.5 W for 190 s = 56,525 J; output computed as 8 mL evaporated x 2260 J/g + 600 g x 15 K x 4.18 = 55,700 J; Naudin reports the ratio as 0.985 and concludes the method "is reliable". Note for reviewers: the calibration is with a heater, not with the plasma running; the page writes the ratio as "56525 / 55700 = 0.985" (the numbers as written are input/output, 1.015; output/input is 0.985). - Low/medium power
cfr/html/cfrtpwr.htm(2003-05-21) [web]: Tiny-CFR, 150 mL 0.2 M K2CO3, W cathode 2 mm x 25 mm. Low power (glow discharge only): 148.4 W for 499.2 s = 74,094 J in; 26 mL evaporated plus 23 K rise gives 73,181 J out; ratio 0.988 (page wording). Medium power (incandescent cathode): 289.2 W for 82 s = 23,714 J in; 10 mL evaporated plus 21 K rise gives 35,767 J out; ratio 1.51, "power gain 147 W". Naudin's principle, in his words: above about 80-85 C, with voltage over 140 V and enough current density, the cathode becomes incandescent and "Only at this moment, the heat produced is greater than the electrical power input"; he attributes it to atomic hydrogen trapped and recombining at the tungsten surface ("a real cold fusion effect"). Measuring method: Fluke 123 scope for voltage, CIE CA-60A current clamp, type K probe; mass loss by weighing. - CFR v3.3 with a new filtered power supply
cfr/html/cfr33.htm(to 2003-06-05) [web]: W rod 3 mm x 30 mm, stainless grid anode, tests at 182, 212 and 259 V DC; scope current checked against an analog ammeter and a clamp. "the DC Voltage has a major role in the efficiency". Numbers are in images. - The CFR index links many more sub-pages not yet read (cfr10-cfr14, cfr20, cfr30, cfr40/40b/40c/40d, cfrmhmc, cfrnewps,
cfrpd, cfrppcglv, cfrstir, cfrtiny, cfrtiny2, hpcfr2, hpcfr21, democfr, the "ape" and "aehodme" sub-projects, and
files/MizunoTanomalouse.pdf). Next run: fetch them, especially
cfrstir.htm(stirring, which Earthtech found matters) and the CFR v4.0 pages.
2. MAHG (claim 5): run conditions and blank tests
- Genesis page
mahg/mahg1.htm(2004-03-01/02) [web] http://jlnlabs.online.fr/mahg/mahg1.htm: Nicholas Moller "of Spectrum Investments Ltd" contracted Faraday Lab Ltd (Alexander Frolov, St Petersburg) to test Langmuir's atomic-hydrogen recombination as a heat source. Frolov's reported efficiencies: about 130% over most of the range, 124% with rectified DC pulses, 82% when the same 100 Hz pulses were smoothed by a capacitor, about 250% with half-period diode pulses (with the remark that this "gave reason to think that a possible mistake had been made or about a physical mechanism"). Main conclusion quoted: pulsed heating of a tungsten filament at 500-1500 K in hydrogen shows high efficiency. - Tests index
mahg/tests/index.htm(2005-05-23 / 06-02) [web] http://jlnlabs.online.fr/mahg/tests/index.htm: tube water-cooled, hydrogen at 0.1 atm, tungsten filament 0.25 mm. Input measured three ways (analog meters with shunt, Fluke 189 with VC601 clamp, Metrix MTX3352-C scope with clamp), agreeing within +/-2%. Input logged as "Rms Voltage and Rms Current"; output = flow (m3/h) x dT x 1161.11 W. Runs 11-37 pulsed (half-sine DC pulses, 50% duty cycle): average efficiency 224%, net gain 464 W. Runs 38-40 with pure DC: 64% (Naudin: "BLANK TESTS", filament only 774 K, below the 2000 K needed for dissociation; 36% losses). Filament temperature from T = (P/(k S))^(1/4) with S = 1e-3 m2 (Stefan- Boltzmann estimate), Langmuir's 1926 dissociation curve; atomic hydrogen 1-7% produced "during only 25 ms each 100 ms". Naudin states the results confirm Frolov's. - Long runs
mahg/tests/lrtests.htm(2005-06-02) [web]: 30 min, 1 h and 2 h at full power; about 875 W output, about 240% efficiency, net 515 W; the dielectric ring runs above 60 C and about 30% of losses go through it (blank tests). - High-efficiency runs
mahg/tests/pultests.htm(2005-06-20/23) [web]: supply changed to a 12 V 100 Ah battery through a MOSFET switch driven by a TGP110 pulse generator; "efficiency ... increases ( up to 1153% ) while the duty cycle ... decreases (up to 5%)"; 2 h runs at 10% duty cycle at about 10 W input. Input is "average voltage and current ... measured on the digital oscilloscope" (the page does not say whether power was averaged as v x i or as product of averages). - Radiation check
mahg/tests/mahg2e.htm(2005-07-08) [web]: Gamma-Scout Geiger counter beside the chamber during run 87 (1 h): "no anomalous radiations". - The logbook page lists Moller as source, with permission.
How it bears: claim 5's numbers can now be checked against the stated method. The reviewers should look at (a) the RMS x RMS input definition for half-sine pulses into a non-linear filament, (b) the pure-DC "blank" at 64% versus pulsed runs, (c) the change of supply and of input-measurement method between the 224-240% runs and the 1153% runs. Frolov's own smoothed-DC 82% versus pulsed 124-250% is the same pattern in an earlier, separate set-up.
3. MEG (claim 3)
meg/meg.htm(project index, jnaudin.free.fr) [web] http://jnaudin.free.fr/meg/meg.htm: links pages not in the folder:criticofsb.htm(critique of Bishop),megjln01.htm,megnot01.htm,megcs01/02.htm,megdiag.htm,megdsqth.htm,meg4cf.htm,magtrans.htm, and prior-art patents Naudin lists (FR2312135, US6362718, WO0007285, CA2172240, DE3501076, US2802170, US2883604, US3087108, US3239771/2, US3368141, US3815030, and Delphion US4006401, US4077001, US4675615, US5926083), plus arXiv physics/0207027 and the Foundations of Physics Letters 2001 MEG paper.meg/megjln01.htm(2000-11-03) [web]: Naudin's "Magnetic Amplifier Experiment": adding a magnet to one side of a ferrite rod raised output on 100 ohm from 13.78 mW to 30.39 mW while input went 853.2 to 856.2 mW. Naudin: "There is NO Overunity here, but only some food for thinking". This is his own statement of the MEG principle (magnet flux switched, not supplied, by the input).meg/megnot01.htm(Jon Flickinger, posted to the jlnlabs group) [web]: Flickinger says COP > 1 is seen only with non-linear resistive loads whose resistance falls with voltage (e.g. varistors), not with linear resistors; "Those of you powering up your MEG for the first time with pure resistive loads, will find the waveforms do not match Bearden's nor JLN's!" This bears on claim 3: the load type is part of the claimed effect (Naudin's own MEG v3.0 note, quoted by Bishop, says a "conditionned" 100 kohm load or a MOV is "REQUIRED").- MEG v3.0 pages (megv3.htm etc.) are gone from both domains (404). Wayback for jlnlabs was "Temporarily Offline" to scripts; next run: try Wayback in Chrome for jnaudin.free.fr/meg/megv3*.htm.