Ramsey Fringes: Precision vs Interrogation Time

Interrogation time T

Free-evolution time T 1.00 ms

Coherence decay time T2

Dephasing time T2 1.20 ms

Probe detuning Δ

Microwave detuning Δ 0.0 rad/ms

Live readout

Fringe width (FWHM)

–

Fringe contrast

–

Precision at this T

–

Best possible (T=T2)

–

Balanced: fringe is reasonably narrow and coherence hasn't decayed too far yet.

How this trade-off works

Two short π/2 microwave pulses bracket a free-evolution gap of length T. During that gap the atom's superposition phase winds up at a rate set by the detuning Δ between the microwave and the true atomic transition frequency. Scanning Δ and recording the final excited-state population traces an oscillating Ramsey fringe:

P(Δ) = ½ · [1 + C(T)·cos(Δ·T)]
C(T) = exp(−T / T2)

Longer T narrows the central fringe (better raw frequency resolution) but gives environmental noise more time to dephase the atom, shrinking the contrast C(T) — the fringe washes toward a flat line.
Shorter T keeps contrast high, but the fringe is so wide that the detuning can't be pinned down precisely.
The achievable frequency uncertainty scales as 1 / (T·C(T)·√N). Because T·exp(−T/T2) peaks exactly at T = T2, that is the optimal interrogation time — the sweet spot atomic-clock designers aim for.

Ramsey fringe — population vs detuning ● theory ● current Δ

Achievable frequency precision vs T (lower is better) ● current T ┊ optimum