Atomic Clocks

last updated 2026-05-04

Physics / mechanism

Atomic clocks exploit the quantised transition frequency of an atom as a reference oscillator. A local oscillator (typically a quartz TCXO or VCXO) is locked to an atomic resonance via feedback: the interrogation signal is swept, the transition signal is detected (optically or via microwave absorption), and any frequency error corrects the oscillator. Caesium beam clocks define the SI second at ~9.192 GHz; commercial Cs standards reach 1×10⁻¹³ fractional frequency accuracy. Rubidium cell oscillators (10⁻¹¹–10⁻¹² accuracy) dominate size/cost-constrained applications. Optical lattice clocks push to 10⁻¹⁸ instability but remain lab-bound. Chip-scale atomic clocks (CSACs) using coherent population trapping (CPT) in Rb vapour cells achieve ~10⁻¹¹/day in <17 cm³ at ~120 mW—driven by Microsemi/Microchip, Safran, and SiTime’s MEMS-enhanced variants.

Competitive landscape

Rubidium VCXOs and oven-controlled XOs (OCXOs) compete below the CSAC tier on cost but lose badly on holdover under GPS-denied conditions. MEMS oscillators (SiTime, Resonant) close the gap in phase noise but not in absolute accuracy. Hydrogen masers outperform Cs in short-term stability (10⁻¹⁵ at 1 s) but are rack-sized. Optical atomic clocks (Sr, Yb, Al⁺) are the long-run successor technology.

Companies using

Connected ideas

Sources

Frontier (open questions)

Frontier questions