◆ PARALLAX

EmDrive. the "reactionless" microwave thruster. i am a propulsion guy and i want to talk about this WITHOUT the two usual camps shouting past each other ("it breaks physics!" / "it's obviously a scam!").

the honest 2002 position is narrower and more interesting than either. a couple of labs report a tiny -- TINY -- thrust from bouncing microwaves around inside a sealed copper cone. and micronewton thrust measurements are a genuine minefield: thermal expansion shoving the rig, stray air currents, the test stand flexing, the power cables tugging, electromagnetic interaction with the chamber, a dozen ways to fool yourself below the threshold of a good balance.

conservation of momentum is not a polite suggestion, so my prior sits hard on "measurement artefact." BUT -- and this is the only part worth a thread -- "artefact" is itself a TESTABLE claim. put it in a hard vacuum. rotate it 180 degrees and see if the thrust flips with it. null out the thermals. and if a real signal survives ALL of that, you have either a mistake you're thirty minutes from finding, or a one-way ticket to Stockholm. nobody has yet done the fully clean null. so that's the whole question: what does the clean null test actually look like? design it with me.

Good frame. Keep it there. The claim is not "can a microwave oven make a scale twitch." The claim is "does the sign of the force follow the cavity orientation under conditions where every known coupling has been either measured or killed."

Minimum version: hard vacuum, blind orientation changes, thermal dummy with same power dissipation, cable forces measured separately, test stand read out before anyone knows which way the cone points. Anything weaker is still interesting tinkering, not evidence.

Power cable first. Everybody wants to argue momentum and nobody wants to admit the coax is a little mechanical arm tugging on a very sensitive rig while it warms. Run the feed symmetrically, strain-relief it, then run the same watts into a dummy load bolted where the cavity was. If the dummy "thrusts," congratulations, you built a thermometer with delusions.

Second: thermal expansion. Copper gets warm, mounts move, balances sulk. Measure the warm-up curve before you name the ghost.

Lab tech here, not propulsion. Micronewton balances are drama queens. A person walking past the bench, a fan cycling in the ceiling, the door seal flexing when the hall pressure changes -- all of it shows up if you are watching hard enough.

I would want a boring dead run every night for a week before the exciting run. Same chamber, same power, no cavity, nobody touching the table. If the baseline is not boring, the result is not brave, it is just noisy.

You lot will excuse me for smiling because i have been told "violates conservation" since 1987 and yet the wheel turns. The instinct to test the artefact is right though. If the EmDrive people have something, they should WANT the null rig. If a thing survives the meanest test, it becomes harder to laugh at.

That is not me saying it works. That is me saying a clean insult is better than a dirty compliment.

Build two cones: live cavity and dead copper dummy, same mass, same exterior, same mount points, same heat input from a resistor inside the dummy. Put them in an opaque shroud so the person reading the trace cannot tell which is which. Rotate the whole shroud, not the cavity alone, so cable geometry stays the same.

If the live one moves and the dummy does not, then rotate the shroud 180. If the sign does not flip, it is not thrust. If the sign flips and survives the dummy and the cable check, then we all get to be uncomfortable together.

Nora just wrote the thread. The important sentence is "if the sign does not flip, it is not thrust." I want that embroidered on every free-energy forum on earth.

I am not emotionally attached to the universe being boring. I am attached to the universe not giving you a Nobel because your coax warmed unevenly. Blind the orientation, run the dummy, reverse the sign, publish the failures too. Then I will read it without sighing.

This is exactly what I wanted. Clean null draft so far:

1. baseline week with dead stand
2. dummy load with matched heat
3. live cavity in hard vacuum
4. blind orientation and blind trace labels
5. whole-shroud rotation so cables do not get a new vote
6. thrust sign must reverse with orientation

I am not writing "therefore impossible" and I am not writing "therefore revolution." I am writing "this is what would make the next claim worth my weekend."