A physics engine, and a verification tool enforced on top of it. PhysWall inverts a closed, published, non-linear physical law at a single measurement point — and refuses when the inverse is not unique. Seven laws, one engine, and the same refusal in all of them.
Each one states its own maturity, and refuses when the measurement cannot carry a conclusion.
The quantity a matching network is limited in is impedance bandwidth — how wide a band a passive network can hold to a return-loss target. Bode 1945 and Fano 1950 put a ceiling on it, and no effort on the network moves that ceiling.
Before asking whether a target is reachable, it is worth asking how well the measurement behind it is known — a bound met by a number with a wide band is not met.
There is no single label for the app. Five are beta, two are filters, and the difference is a different job rather than a lower grade.
The first of them is the Bode-Fano limit, and it is the oldest: Bode bounded the matching a single reactive load allows, and Fano generalised it to any reactive load. Steer states the four inequalities in Microwave and RF Design III, chapter 7.2, which is free to read. What they say is that a matching network has a budget, and spending it on bandwidth means spending less on depth of match. Run backwards, the same inequality answers a question an engineer actually has: no network reaches that return loss over that bandwidth, and the two weeks are better spent elsewhere.
D1 Bode-Fano antenna matching beta D2 f* = kT/h thermal ceiling beta D3 conductor loss PCB T-critical beta D4 photonics BF nanoparticle bound filter D5 NTK-BF cascade three literatures filter D6 Landauer bit erasure beta D7 anapole disk radius beta
D4 and D5 cannot be inverted, and that is arithmetic rather than a gap in the work. The polarisability is a system, not one expression to solve back; the cascade has three unknowns against one reading. A filter answers "is this even possible", completely and correctly. It does not derive, and it says so instead of returning one of the many answers that fit.
The full interface is a single HTML file. No build, no dependencies, no network — measurements stay on your machine, which is the point rather than a convenience.
A single HTML file. No build, no dependencies, no network — your measurements stay on your machine.
It will not return a number when no load can meet the spec. It will not give a point estimate where the measurement only supports a range. And where two sources disagree about whether something is resolvable at all, it declines rather than picking the one that suits.
PhysWall was developed and architected by Gadi Zion.
Built on PhysWall — the same engine reads antenna bandwidth, conductor loss, bit erasure and heat limits. It answers what the measurement implies, and refuses when the measurement cannot say.⚠ Check this instead of believing it. Every number here reproduces from a source that is named, and the claims that turned out wrong are still printed next to what replaced them. The same engine runs all of these — it asks how much a measurement allows you to conclude, and refuses the same way in every field. The same engine runs all of these — it asks how much a measurement allows you to conclude, and refuses the same way in every field. How to check each one →