An axion haloscope (ADMX-style) places a tunable microwave cavity inside a strong static magnetic field B. If dark-matter axions exist, the inverse Primakoff effect lets the virtual field convert an axion into a real microwave photon only when the cavity's resonant frequency matches the axion's Compton frequency:
E = m_a c² = h f_c
m_a[μeV] ≈ 4.1357 × f_c[GHz]
The cavity supports a TM₀₁₀ mode whose axial electric field follows the Bessel profile E_z(r) ∝ J₀(x₀₁ r/R), x₀₁ ≈ 2.405 — rendered below as a side cross-section, radial position left↔right, cavity height top↔bottom. A movable tuning rod perturbs this mode and shifts f_c as it sweeps from the cavity's center toward its wall.
The expected conversion power follows the standard haloscope formula (Sikivie 1983; as used by ADMX):
P_sig ∝ g_aγγ² · B² · V · C · Q_L · ρ_a / m_a
Doubling B quadruples the signal; a higher loaded quality factor Q_L both boosts power and narrows the cavity linewidth Δf = f_c/Q_L that must be scanned bin-by-bin. The thermal/quantum noise floor follows the Dicke radiometer relation P_noise = k_B T_sys Δf, so SNR = P_sig/P_noise for a 1-second dwell. A real experiment steps the rod slowly across the whole band, integrating longer at each point, because a true signal only appears in the narrow bin where f_c happens to equal the (unknown) axion mass — reproduced here as the hidden target frequency you can reseed and hunt for.
This 2D version renders the cavity as a flat side cross-section instead of a rotating 3D model, so the field pattern, tuning rod and solenoid arrows are all visible at a glance without needing to orbit the camera.