The reader antenna is modelled as the square trace a real PN532 board actually uses (not a circle) — the red sprite is the board itself, sized and shaped to match. The donut is an iso-surface: it traces the boundary where coupling equals your read threshold. If it looks like a filled-in blob rather than a ring with a visible hole, that's not a bug in the physics — it means even the weakest interior spot (the dead centre of the loop, farthest from the copper) still clears your threshold. Raise the threshold (or lower the power) until a hole actually opens up: that hole is the true weak spot, and it sits at the geometric centre — coupling is usually strongest right at the trace itself, not in the middle of the board. So placing the tag flush and centred works fine at healthy power, but if you're marginal, nudging it toward the copper outline rather than dead-centre can help. The thin lines threading through the donut are the coil's actual magnetic field lines (Biot–Savart around the square loop) — the same closed loops, split by colour where they cross to the housing side (amber) vs. the tag side (cyan). The badge next to the tag is your live read estimate — drag the tag around until it goes green. Turn on 2nd reader to test two PN532 modules facing each other: the field lines and both donuts are recomputed from the two coils' combined field, so you can see where they cancel or reinforce as you change the spacing between them. A second chip also appears, always exactly 58.65 mm from the first along the core — a fixed physical distance that does not stretch when you move reader 2 — each with its own live read badge, so you can tell whether one, both, or neither is likely to read.