Gray-EV: Science Contrarian review

Reviewer role: Science Contrarian (PhD-standard; judges by experiment, applies the same rigour to Gray and to his critics). Source folder read: source folder: Gray-EV (read only).

Source gap (repeated from Step 1). The four patents are scanned images with no text layer. I read the whole of US 3,890,548 p.1 (abstract) and the background page (printed pp.1-2), plus the full GB 2,030,801 specification page, and the Crosby letter page image. The middle description pages of US 3,890,548 and most of US 4,595,975 and US 4,661,747 were not read by me or by Step 1. The press articles have poor OCR. There is no independent test report, no waveform, no instrument list and no data file in the folder. Where a conclusion below rests on [memory] alone, it is marked.

Web verification: my web-search budget was exhausted by the session limit, so only one web call (Google Patents, US 3,890,548) succeeded. Every other medium/low-confidence [memory] fact is flagged "web check failed".

Corrections to Step 1

1. The Crosby letter does NOT say efficiency "exceeds 100%". The page image reads: "The measured overall system efficiency exceeds 99%" [read: page image, crosby_letter.pdf]. Step 1's OCR guess ("ggt" = 100%) and its claim 2 wording ("measured efficiency exceeded 100 percent") are wrong. This matters, because the letter is then internally inconsistent (see Claim 2), instead of simply asserting over-unity.
2. The letter's figure is 26.8 watts per hour, not 26.9 [read: page image]. The text is "The total battery power consumed by the motor during a 21 minute run was only 9.75 watts; this equals 26.8 watts per hour." (The "26.9" in Step 1 may come from the garbled TATTLER reprint; I could not verify that.) The two numbers are also not quite consistent: 9.75 Wh in 21 min (0.35 h) is 27.9 W average; 26.8 W over 0.35 h would be 9.38 Wh. Small, but sloppy for a "test report".
3. Battery capacity is "5,454 watts for one hour" (i.e. 5,454 Wh), confirmed [read]. The signatory is J. A. Maize, Director of Engineering, and the page carries a notary seal (Carol L. Dye, 9/28/73). A notary attests only the signature, not the test. The letter is addressed to "Larry Crosby" at the institute's own address. Step 1's statement that the institute is "named after the financier's family" is not in the letter; treat as unverified [memory] low, web check failed.
4. static_generator.pdf is identifiable. It is a rotated one-page EvGray product sheet, "EVGRAY STATIC GENERATORS" [read: page image]. The "Generator-Only" model gives about 45 W from a 12 V car battery to run standard 110 V lights, radios, TVs; the "Generator-Battery" model gives about 30 W for about 6 h from a self-contained rechargeable battery, priced about $50 and $75. It states "No moving parts". So "static" here means solid-state (no moving parts), i.e. an inverter, like ems_inverter.pdf. It is not a static-electricity device. This undercuts Step 1's guess that it relates to atmospheric static electricity. A 30 W x 6 h = 180 Wh battery is an ordinary figure; nothing over-unity is claimed on this sheet.
5. Step 1 missed a simple check on the Crosby numbers: a 21-minute run at 7,460 W is 2,611 Wh, about 48 percent of the whole 5,454 Wh bank. So the test was physically possible on battery power alone, with no anomaly, if the real draw were ~7.5 kW. The "anomaly" exists only in the claimed meter reading of 9.75 Wh (see Claim 2).
6. Confirmed: US 3,890,548 filed Nov 2 1973, granted Jun 17 1975, inventor Edwin V. Gray [web: https://patents.google.com/patent/US3890548A/en]. The patent text says the engine "is believed to operate on the principle of conservation of energy" and the "discharge overshoot or back e.m.f." charges "a secondary battery 10a to store this excess energy" [read: p.1-2 page image; web, same URL]. Step 1's reading that the patent itself makes no over-unity claim is correct.
7. Step 1 tagged its "pulsed-input mismeasurement" explanation [memory] medium. Web check failed (budget). It stays [memory] medium. It is a hypothesis; no instrument details exist to confirm it.

Claim 1: Pulsed-capacitor-discharge motor (US 3,890,548)

a. Mathematics. The patent has no equations. Its mechanism (same-polarity cores repel; each rotor/stator pair is fired by a spark gap when aligned; rotor coasts "on the inertia of the repulsion action"; capacitors recharge between pulses) is dimensionally and logically consistent. The statement "current is needed only at the instant of a capacitor discharge" is true of the circuit but says nothing about energy: each discharge costs the energy ½CV² taken from the source, whatever the duty cycle. The patent does not mistake this; it says "conservation of energy" [read].

b. Established experiments. Repulsion pulse motors, spark-gap switching, capacitor-discharge coil guns and ignition systems are all standard physics and engineering [memory] high. Nothing in the described device contradicts a known measurement. Spark gaps are lossy switches (arc and ringing), so efficiency would be expected to be modest, not extraordinary [memory] medium, web check failed.

c. Tested? The patent was granted (an examiner found no prior art; the patent text says the applicant's search found none). Granting does not test performance. Step 1 documents no neutral test of the motor. The AP item reports TRW scientist Donald Cronin saw a test "about a year" before July 1975 where the motor "ran for about 10 seconds and then everything stopped" and was not convincing [read: various_articles.pdf, garbled OCR]. That is an informal observation, not a controlled test, with no measurement. A Japanese group reportedly had equipment but left without signing [read, same article]. No replication on record.

d. If true. As written (a conventional if inefficient pulse motor) nothing in mainstream physics changes.

Verdict: Plausible, untested. It works as a mechanism in principle; no data exist on its output, efficiency or sustained running.

Claim 2: 10 hp for about 26.9 W of battery drain, efficiency over 100%, 203 h runtime (Crosby letter, Sept 1973)

a. Mathematics. I checked each number against the page image. - Torque 48 ft-lb = 65.1 N m; speed 1,100 rpm = 115.2 rad/s; P = 65.1 x 115.2 = 7,497 W. The letter's 10 hp = 7,460 W is consistent (7,457 W for 10 mechanical hp) [read + calculation]. - 5,454 Wh / 26.8 W = 203.5 h, matching "203 hours". Consistent arithmetic. - 7,460 W / 26.8 W = 278. That is an "efficiency" of about 27,800 percent. - But the same letter says "efficiency exceeds 99%". An efficiency of 99 percent with 7,460 W output needs about 7,500 W input, which drains 5,454 Wh in 0.73 h (44 min). So the letter contradicts itself: either the efficiency is about 99 percent (and the 203 h figure is false) or the input is 26.8 W (and the efficiency is hundreds of times 100 percent). The letter reconciles them only with "electrostrictive power ... developed" and battery energy "recycled", which are not defined or quantified. Electrostriction (strain proportional to E squared in a dielectric) converts energy; it does not create it [memory] high. - 9.75 Wh in 21 min is 27.9 W, not 26.8 W (minor).

b. Established experiments. If battery chemical energy is the only source, output energy cannot exceed it. Over 21 minutes the claimed output is 2,611 Wh versus a claimed drain of 9.75 Wh. That ratio would violate conservation of energy, which is among the best-tested results in physics. The simpler candidates are measurement errors: a DC ammeter or voltmeter reading the average of a high-peak pulsed current, or battery voltage sag/recovery misread as capacity. These candidates are [memory] medium, web check failed, and the letter gives no instruments, wiring, sampling or waveforms, so I cannot confirm or exclude them.

c. Tested? The letter says "extensive testing" by "Pan World Enterprises Co., Ltd., and EvGray engineers" (May 1-10, 1973), "confirmed" Sept 25, 1973 by the undersigned and EvGray engineers, and "Tests reports supporting the above information is available at EvGray." So the testers were the promoters themselves or their associates. No neutral lab, no instrumentation, no report in the folder. Not replicated. The letter is not independent and not a test report; it is a summary letter. Was it dismissed without testing? Partly: the deputy DA said the device "would violate all the laws of physics" [read: AP item], a prior-based dismissal; the only recorded observation by a technical outsider (Cronin) was a 10-second run. Neither is a controlled test of the 26.8 W claim. A decisive test was easy and cheap (a calibrated power analyser or a calorimetric/charge-balance test on the battery), and its absence over 1973-76, while investors were being solicited, is itself evidence against.

d. If true. Conservation of energy would fail in a macroscopic, easily repeatable circuit. Energy accounting in electrical engineering (and the Poynting theorem) would need a new, undetected source. A 278:1 effect is not subtle; it would be reproducible by any competent lab.

Strongest honest case for Gray. Pulsed-current power is genuinely hard to measure with DC meters, and mismeasurement is a real, historically documented error class [memory] medium. A well-run test could have exposed it. Gray did not get one, and a TRW scientist's 10-second observation is consistent with the machine not sustaining operation. Honest case against: the letter's own 99 percent versus 27,800 percent contradiction, the unusable "5,454 watts for one hour" phrasing, the "watts per hour" unit error (watts per hour is not a power; the writer means watts), and the absence of data.

Verdict (split): - As a measurement report: Internally inconsistent (99 percent efficiency versus 26.8 W for 7,460 W; "watts per hour" unit misuse). - As a physical claim (278:1 output/input from batteries alone): Refuted by calculation (energy conservation: 2,611 Wh out in 21 min cannot come from a 9.75 Wh drain). - As to the actual machine's real performance: Never properly tested (no neutral, controlled test; not dismissed without testing outright, since a 10-second run was watched, but no measurement was made).

Claim 3: Back-EMF recovery recycles most of the energy (GB 2,030,801; US 4,595,975; US 4,661,747)

a. Mathematics. I read the GB specification in full. Phase A: 24 V boosted to about 2,000 V charges the capacitor. Phase B: switch 6 closes; capacitor discharges through switch 1, the inductive load and into recovery battery 2. Phase C: when capacitor voltage reaches zero, "the inertia of the current in the inductance" would reverse-charge it; a diode path instead keeps the current flowing in a closed loop through the load into battery 2. This is the standard freewheel/flyback path. The inductor's stored energy ½LI² is delivered to battery 2 (minus I²R and diode losses). Consistent. The patent's claim is merely "a more efficient use of energy" and its sole claims are "substantially as hereinbefore described" [read: page image]. No quantitative claim at all. US 4,595,975 and 4,661,747 repeat the idea (second capacitor stores back EMF and returns it to the source) with a "conversion switching element tube"; I read only abstracts and claims pages via Step 1, so I could not check their internal consistency.

b. Established experiments. Flyback diodes, regenerative braking, switched-mode converters and resonant/charge-recovery drivers are everywhere [memory] high. Recovery of the reactive (magnetic) energy of a coil is well established. But energy delivered as mechanical work to the rotor is not "back-EMF" and cannot be recovered this way; only the coil's reactive energy can.

c. Tested? No test data in the folder. Whether Gray's circuits recovered any measured fraction is unknown. The circuit principle itself is tested by decades of power electronics [memory] high.

d. If true. Weak form (a diode path returns some inductive energy to a battery): nothing changes. Strong form (recycling gives net output exceeding input, the promoters' claim): conservation of energy would fail.

Verdict (split): - Weak form (partial recovery of coil reactive energy): Established in principle; the fraction achieved in Gray's machine is unmeasured. - Strong form (recycling making the machine self-sustaining or over-unity): Refuted by calculation (recovered energy is at most the stored reactive energy, which came from the source; the output work is additional and is never returned) and never supported by a measured test. The patents themselves do not make the strong claim.

Claim 4: "Cold" or static electricity, possibly atmospheric, rather than ordinary battery current

a. Mathematics. No equation or definition exists. The sources are the promotional article (a magnet "flew off" and was caught bare-handed; "cold" electricity; speculation about ball lightning and atmospheric static) [read: ems_electronic_power.pdf, text layer]. The patents do not use the term. Per correction 4 above, "static" in EvGray product literature means no moving parts (solid state), not static charge.

b. Established experiments. A 3,000 V pulse from a small capacitor has a small energy; a 1.25 lb magnet thrown by a brief repulsion pulse and a cool-running coil are explained by low average power (low duty cycle), not a new energy form [memory] medium, web check failed. Atmospheric electricity is a real, measurable potential gradient (~100 V/m fair weather) but its extractable power density is tiny; no claim is tested [memory] medium, web check failed.

c. Tested? Never tested in any way I can find. No instrument, no measurement of "coolness", no antenna or earth connection described. Gray's own sense of "cold" electricity came from the radar/lightning analogy [read: article].

d. If true. A new form of electrical energy distinct from ordinary charge and current would be required, or a practical atmospheric energy tap.

Verdict: Not testable as stated ("cold electricity" is undefined). The sub-claim "taps atmospheric static electricity" is Never properly tested; it is not in the patents and was not dismissed by anyone with a test.

Claim 5: The LA DA raid and prosecution (1974-76) were meant to suppress the invention

a. Mathematics. Not applicable; the logic of the claim: an allegation of motive.

b. Documented facts [read]: The tabloid says a July 1974 raid seized records, plans and the latest prototype; no charges for months; later eight counts of grand theft and others were filed, then dropped, leaving two securities misdemeanours. The AP item of Jul 7 1975 reports Gray, 50, charged with bilking investors; Deputy DA Mitchell Harris said the device "would violate all the laws of physics"; the DA's office said the delay was because investigators were busy and prosecutors found few willing investors; charges were filed in May. The 1976 magazine says "In March, 1976, Gray pleaded guilty" [garbled OCR, read] and paid a fine. Gray raised about $2 million from 788 stockholders [read: promoter-side article]. Those facts fit an ordinary investor-protection case. They also fit suppression; they do not discriminate.

c. Tested / how it was rejected. The DA never tested the machine; the Harris quote shows a prior-based judgment about physics by a prosecutor, which is not a reason to charge fraud (fraud depends on what was told to investors). That the eight counts were dropped is reported only by the tabloid (partisan); I have no court record [web check failed]. The guilty plea is a point against the suppression thesis, although plea decisions are not evidence of guilt or innocence by themselves and I have no records of its terms. Funding was damaged (loss of Denver backers, plastic-car project dropped) [read]. No evidence of blocked publication or reanalysed data exists in the folder. The "Detroit/oil conspiracy" parallels in the TATTLER piece are asserted by analogy only.

d. If true. Nothing in physics; it concerns sociology of science and law. It matters for the evidence only if the motor really worked; it cannot show that it did.

Verdict (split): - The raid, charges and plea occurred as reported by the articles: documented in the sources read, with tabloid-grade reliability for details; court records not seen. - The claim that the purpose was to suppress the invention: Not testable as stated (motive) and unsupported by the documents read. Whether the device was properly tested before or after is covered under Claim 2: Never properly tested.

Summary table

Fairness note on the critics

The deputy DA's "violates all the laws of physics" is an appeal to theory, not a test; the only technical eyewitness, Cronin, saw a 10-second run and found it unconvincing, which is not a controlled test either. Neither side produced an instrumented, neutral measurement; the burden lies with Gray, who had the device and the funding to have one done. The single experiment that would settle claim 2: measure battery input with a power analyser or calorimetric/charge-balance method (integrate V x I at kHz-MHz bandwidth) during a dynamometer run of the same machine, logged by an independent lab.