This is the same NEMS resonant mass sensor as the 3D beam-bending version, but read out the way a real spectrum/network analyzer actually diagnoses one: not the physical shape of the vibrating beam, but its steady-state frequency-response curve — the analytic solution of a driven damped harmonic oscillator, computed independently of the 3D engine and swept across a frequency window every frame:
ω₀ = √(k/m) f₀ = ω₀/2π
A(ω) = (F₀/m) / √[(ω₀²−ω²)² + (ωω₀/Q)²]
Δf/f₀ ≈ −Δm/(2m) Δf(FWHM) ≈ f₀/Q
Each time a particle lands, the total mass m increases by Δm, and the whole resonance peak steps to a lower frequency. The previous curve is kept on screen as a fading trace, so the sequence of landings builds a waterfall of stacked curves — the leftward staircase march of the peak is the same signal a real NEMS lab plots to weigh individual viruses and molecules, one landing at a time.
- Quality factor Q controls the curve's sharpness: a higher Q gives a narrower, taller peak (better frequency resolution, hence better mass resolution) but takes the physical device longer to ring down after a change.
- Linewidth Δf (FWHM) is the width of the peak at half its maximum height — it sets the smallest frequency shift, and therefore the smallest single-particle mass, the sensor can reliably distinguish from noise.
Real-world relevance: this frequency-domain view is exactly what a lock-in amplifier or network analyzer shows when characterizing a real NEMS/MEMS resonator — the mechanical picture and this spectral picture are two readouts of the identical physics.