RankingsPrimerArbiterPhysicistMathematicianContrarianQESites logClaimsSourcesResearch notes

Science Contrarian review: Radiant_One_Youtube

Reviewer: Science Contrarian (Step 2). Standard: experimental evidence over consensus; strongest honest case for the claims; the same rigour applied to the inventor. Tags: [read] = from the project file or the Investigator's saved notes; [web] = URL given; [memory] = recalled, confidence given; "calc" = my own arithmetic from numbers cited.

Source gap (repeated from claims.md)

The project folder holds one image only: a schematic screenshot, "Ground Energy Extraction (GndNrgEx) Circuit 1.1 (Corrected)", channel radiant_one [read: page image p.1]. No deep/ reads exist (none were needed for a one-page image) and research/files/ is empty. Everything about measurements comes from the Investigator's notes of YouTube descriptions and auto-generated transcripts [web, via research/2026-10-04-youtube.md]. I did not re-watch the videos (web budget not spent on that; the Investigator's notes quote them closely). Conclusions beyond circuit analysis rest on those notes, and the Investigator could not read GrnNrgEx_5, GndNrgEx_1 or _10 transcripts. Where I rely on [memory] I say so.

Corrections to Step 1

I viewed the PNG myself. 1. Component values in facts.json are correct (V1 11-13 V, R1 1 Meg, C1 470 pF, Lb 1.4 uH, L1 8.6 uH, L2 26.7 uH, Q1 MPSA06, two 1N4148 with a clip tap, two 26-LED series strings each with an anti-parallel 1N4148, Earth Rod) [read: page image p.1]. The digest's "41.13 L" and "26.71" are OCR garbage, as Step 1 says. 2. facts.json calls the orange bar "apparently an overlay marking the coupling between L1 and L2". I see it as two orange vertical blocks above and below the L2 line that look like a video-editor mask or redaction over the middle of the schematic, cutting the title text ("Ex[]action"). It is not a labelled element. Its meaning is unknown; do not read it as a coupling symbol. The circuit as drawn shows no explicit mutual-inductance marking between L1 and L2; L2 sits in series with the collector node, between the collector and the LED load [read: p.1]. 3. Topology correction: L1 is the collector inductor to the supply rail; the collector node also feeds L2 directly. So L2 is a series choke/resonating inductor into the load, not a transformer secondary as the bar might suggest. 4. claims.md claim 2 says the circuit "can drive" the load; that is the video author's observation (lit LEDs), not something visible in the file. Claim 1 and 3 are implied from title and note; the author's own words are more careful (see below) and should be cited next to the implied claim. The Investigator found the author writes in most video descriptions "no overunity performance is suggested or implied" [web: research/2026-10-04-youtube.md], while also framing the work as Moray-style ground-energy extraction and saying a small driver input produces "a much larger response from the ground electrons" (GndNrgEx_10, https://www.youtube.com/watch?v=Bhj8N7_stZY). Both statements are kept; they are in tension and neither is discarded. 5. Step 1's mainstream_conflict says the title "implies extracting net energy" and conflicts with conservation. That overstates: the author's measured numbers (below) do not exceed the supply input. The conflict is with the interpretation (a ground "energy" source), not with an energy-budget violation. 6. Missing from claims.md: the author's heat test and the 4 mW-per-LED figure (Investigator leads), and the "ground energy stops at a 1N4148" principle. I add them below.

What is actually claimed (author's numbers) [web, via Investigator]

Energy budget (calc): input 12 V x 30 mA = 0.36 W (0.39 W at 13 V). LEDs: 52 x 4 mW = 0.208 W, i.e. about 58% of input. Operating frequency of the Lb-C1 branch: f = 1/(2 pi sqrt(Lb C1)) = 1/(2 pi sqrt(1.4e-6 x 470e-12)) = 6.2 MHz (calc; the real oscillation frequency was not measured in anything seen). So on the author's own figures no excess energy is shown. I stress that "4 mW per LED" is itself an unreported-method number: nothing seen shows how RF LED power was measured, and the author himself warns meters read incorrectly in the RF field.

Claim-by-claim

Claim 1: A circuit named "Ground Energy Extraction" can draw usable electrical energy from the earth through an earth rod.

a. Mathematics: There is none; no equations are given. The schematic is internally consistent as a circuit (a SEC-15/Slayer-type single-transistor oscillator feeding an LED load) [read: p.1]. Energy bookkeeping on the author's numbers closes without the ground (0.208 W of LED power from 0.36 W input). b. Consistency with experiment: The observation that a floating RF-driven circuit lights LEDs only when one terminal is connected to earth is entirely familiar in ordinary RF practice: an earth rod acts as a return path (a counterpoise) through the capacitance between the battery-powered circuit and ground. Stiffler's own documentation reportedly calls the earth or long wire "in reality a counterpoise" (Google snippet only, Investigator, not read in full) [web: Scribd "Wireless"]. A rough check (calc): with ~20 pF stray capacitance to ground at 6.2 MHz the return reactance is about 1.3 kohm, so ~1.3 mA of LED current needs only ~1.7 V of RF across that path, easily available at the resonant node. Hence the earthed/unearthed difference does not need a new energy source. It also shows that unearthed LEDs staying dark while the supply current is the same is precisely what a retuned oscillator does: detuned, the RF is dissipated elsewhere (transistor, coils, radiation). Nothing in the record shows ground energy contributing a measured watt. Contrarian caveat: nothing in the record excludes it either, because the experiment as run cannot distinguish "ground supplied some of the LED power" from "supply supplied all of it". c. Tested? By the author only, with a bench supply/battery current meter, an IR thermometer, and visual brightness. No input-versus-output power measurement with an RF-capable instrument, no light-output measurement (GndNrgEx_4 was promised to measure light, and GndNrgEx_4, _6, _7, _8 are absent from the public channel list [web: Investigator]), no controls (e.g. replacing the earth rod with a floating wire or a capacitor of similar value, or earthing through a resistor; running with the earth rod disconnected but the LEDs on a floating plate). No independent replication found; the Investigator found none and one commenter wrote "You didn't show the circuit working" (2018) [web]. Related-community builds of the SEC-15 itself at 12 V, 30-75 mA exist (Energetic Forum, 2011-12) but those do not test the ground claim. d. If true: only the strong form (ground electrons supply net power beyond the supply input) would require physics to change: it would imply a usable power source not accounted by known earth-electric-field/telluric-current magnitudes, which are very small. [memory, medium: fair-weather atmospheric field ~100 V/m, current density ~pA/m^2; I did not web-verify this.] The weak form (earth is a return path that completes the circuit) changes nothing. Dismissed or tested? Neither in any serious sense: the claim was not formally tested by a neutral party, and it was not formally dismissed either; it sits on a small YouTube channel with a few thousand views per video (14,645 views for the schematic video [web: Investigator]). Moray's own claims are a separate record I did not check against sources. Verdict: split. - Weak form (earth rod as RF return/counterpoise lets LEDs light): Plausible, untested (consistent with known RF behaviour; no controlled test seen, easily done). - Strong form (net energy extracted from ground): Never properly tested. The one energy budget available (0.208 W out vs 0.36 W in) neither supports nor refutes it, since it shows no excess. Not dismissed without testing, rather not tested.

Claim 2: The MPSA06 oscillator (values above) coupled by L2 can drive an earth-connected load of two 26-LED strings with anti-parallel 1N4148s.

a. Consistent as a circuit [read: p.1]. A white LED needs ~2.8-3.2 V forward [memory, high], so a 26-LED string needs ~75-80 V of peak RF before conducting; a collector resonance with 12 V supply can reach that (collector swings of several times Vcc are normal for an LC-tuned Class C-ish oscillator). I did not simulate; this is a plausibility argument, not a verified result. The anti-parallel 1N4148 across each string lets the opposite-polarity half-cycle bypass the LEDs, protecting them from reverse voltage. Note the string is a rectifier: the LEDs conduct on one polarity, so current into the earth return is not zero-mean in each string; the DC path closes via the ground-to-supply capacitance, which is a further reason the earth return behaves as a displacement-current path. b. LED lighting from RF-fed, sparsely supplied circuits is well known (Slayer/Joule-thief/SEC builds, Energetic Forum thread) [web: Investigator]. c. Tested: shown on video by the author; no independent reproduction of this exact "Circuit 1.1" found. The supply-current equality depends on the tuning clip, so reproduction would need the clip position. d. If true: nothing changes. Verdict: Plausible, untested (as an independent reproduction; the lit LEDs themselves are not in dispute).

Claim 3: Supply current can be tuned (clip near L2) so earthed and unearthed currents match, which is presented as the operating condition.

a. Not a mathematical claim. Logical point (Contrarian against the inventor): the tuning procedure is "move the clip until the unearthed current equals the earthed current" [read: p.1; web: GrdNrgEx_2]. Equal current is therefore set by the adjustment, not discovered. It cannot, by itself, show that the earth supplies energy. Presenting "equal current" as an anomaly would be circular; the author's own descriptions say "no overunity ... implied", so I do not accuse him of that, but his phrase "The power supply registers no change in the current supplied ... whether the load is connected or not" (GndNrgEx_1 description [web]) will be read as an anomaly by viewers and needs the tuning caveat beside it. b. What it does show, legitimately: that the earthed load (0.2 W as claimed) can be added without increasing supply draw, after retuning. That is ordinary for a load that replaces another loss in a resonant oscillator (the unearthed circuit puts the same power into transistor/RF radiation). The author's heat test (transistor +1.2 C unearthed vs +0.7 C earthed) is a reasonable idea, but weak as designed: a case-temperature rise of one or two degrees after a 5 min run, measured on one small transistor with an IR thermometer that read -50 in the field (unreliable), at which time dissipation elsewhere (coils, wiring, radiation) was not measured. It supports "the unearthed case dissipates more in the transistor" directionally, no more. I note this is a good-faith attempt to answer forum objections (GndNrgEx_3 answers overunity.com questions, 2015 [web]) and should be credited, while its quantitative weakness is stated. c. Tested only by the author. No test under a calibrated RF power meter or a scope with a current probe/series resistor and a bypass capacitor; no calorimetry. d. If true: nothing changes. Verdict: the observation is reported by one source (the author; no second source found) and is not contradicted; its interpretation as evidence for ground energy is Not testable as stated (matching currents is set by the tuning). Never properly tested for the underlying input/output question.

Claim 4: The circuit produces enough RF that multimeters and oscilloscopes can be damaged or read incorrectly.

a. n/a. b. Consistent with standard experience: DMMs rectify/mis-read RF pickup; scope probes in a MHz field pick up with long ground leads [memory, high]. The author reports an IR thermometer reading -50 and personal meter damage [web: Investigator]. c. Not formally tested but not in dispute. d. Nothing changes. Contrarian implication: it cuts both ways. If the meters are unreliable, so is the "4 mW per LED" and the "30 mA" figure unless measured with RF-proof methods (DC current through a bypassed shunt measured with the instrument far from the field, battery-powered recorder). The author's choice of a battery for GndNrgEx_9 is sensible and should be credited: it removes mains-ground coupling and a switch-mode supply as confounders. Verdict: Established (as an experimental caution); it also lowers the confidence of the measured power figures, which are Not properly tested.

Investigator's findings weighed (kept in the record)

  • Twelve-video channel 2015-2018, 469 subscribers; videos on ground energy, single-wire "One Wire Energy" (earth-connected receiver much brighter than unearthed; "Full Metal Jacket" metal-enclosed driver still lights LEDs via a single wire when earthed), and HV arcs [web: research/2026-10-04-youtube.md]. The metal-enclosure test is a sensible partial control against "it is just radio" but does not exclude capacitive/conductive coupling via the single wire and the earth return; it was reported by the author alone.
  • Missing videos _4, _6, _7, _8 (including the promised light-output measurement): competing explanations are never made, deleted/private, or renumbered; the Wayback Machine or the playlist (PLgevW9RSTBFwr5FD3a4V5CY20BYCP18Go) could decide. This is an unexplained gap in the author's measurement record, not evidence of suppression.
  • Lineage: Stiffler SEC-15 exciter, Slayer exciter, Energetic Forum builds at 12 V, 30-75 mA [web: https://www.energeticforum.com/forum/energetic-forum-discussion/renewable-energy/5174-joulethief-sec-exciter-and-variants/page139.html]. Stiffler's earth-as-counterpoise description and radiant_one's "ground electrons" description of the same earth connection are both recorded as their authors' characterizations.
  • Moray (1920s-30s) framing is the author's analogy [memory, high that Moray claimed an unexplained "radiant energy" device and wrote "The Sea of Energy"]; I did not examine the Moray record here and do not judge it.
  • Related YouTube grounding demos (Okki Moeljadi, George Chaniotakis) exist; unexamined.
  • A comment about "Henry Murry", a lab fire and a body in a ditch is one commenter's unchecked story; it does not match the Moray record as known [memory, low]. I keep it as recorded, with no weight.
  • No independent supporting or replicating material, no patent, no company, no publication found. The Investigator's run was short (7.6 of 10 min); overunity.com's 2015 thread, Stiffler's documents, and the remaining transcripts were not reached, so the lack of support is partial absence of search, not evidence of absence.

The strongest honest case for the author

  • Real, repeatable-looking observations: 52 LEDs lit at about 30 mA from a battery, with no change in battery current between connected and disconnected states after tuning, and the observation that the 1N4148 in the earth path changes the response (GndNrgEx_10 description). These are cheap to reproduce, and the earth-return role is a legitimate research question in itself.
  • He publishes component values, coil construction and a tuning method, warns about measurement errors, answers forum objections with a heat test, uses a battery, and repeatedly disclaims overunity. That is a more careful posture than many in the genre.
  • The one unanswered experimental question is clean and answerable: does the earth (versus a floating counterpoise of equal capacitance) change the efficiency of power delivered to the LEDs, and is the delivered power, measured with RF-appropriate instruments, ever larger than the DC input minus all measurable losses?

The case against, applied to the inventor

  • No input/output power measurement with a suitable instrument; "4 mW per LED" has no stated method; equal current is set by tuning; heat test is crude; no control replacing the earth rod with a counterpoise; the series of promised light measurements is missing; the language of "ground electrons" has no definition that could be tested; the claim that a "small energy input produces much larger response" is not supported by any figure above the 58% ratio.
  • "Fibreglass table partly responsible for magnetic-dielectric interaction" (GndNrgEx_9) is an untested, unfalsifiable-as-stated assertion [web].

Experiments that would settle it

  1. Replace the earth rod with (a) a floating metal plate of measured capacitance, (b) a capacitor to the battery negative, (c) the rod via a resistor. If the LED brightness tracks capacitance/coupling, the earth is a return path.
  2. Measure battery V x I with a shunt, a bypass capacitor and a recording DMM away from the field, plus RF power to the LEDs with a calorimeter or photometer; compute efficiency. Energy from the ground would require LED output > input minus all other losses.
  3. Run on the battery alone for hours (42 Ah at 0.36 W is roughly 1400 h [calc: 12 V x 42 Ah = 504 Wh; 504/0.36 = 1400 h]); a long-term comparison of battery discharge with and without earth would directly show any difference.

Summary of verdicts

  • Claim 1: weak form Plausible, untested; strong form Never properly tested (not dismissed without testing; ignored).
  • Claim 2: Plausible, untested.
  • Claim 3: observation reported by the author alone; interpretation Not testable as stated (set by tuning); input/output question Never properly tested.
  • Claim 4: Established as a caution; consequence: the published power figures are unreliable until measured another way.