Real-Time Self-Correcting RF Atomic Magnetometer for Unshielded Operation
ORAL
Abstract
Optically pumped atomic magnetometers (OPAM) offer high sensitivity to rf magnetic fields and are easily tunable with a bias field B aligned with the pump direction. The sensitivity relies on a narrow linewidth, yet a narrow linewidth makes them vulnerable to small ambient fields. For instance, the arrival of a nearby elevator can easily shift the Larmor frequency beyond ~0.5 kHz linewidth, greatly degrading the signal of interest. Therefore, unshielded OPAM use requires real-time autotuning. We demonstrate how atoms themselves can be used to set B, both its magnitude and direction, using a fiber-coupled 87Rb OPAM. The probe beam in this OPAM is perpendicular to the pump and passes through the 0.3 cm^3 atomic volume 35 times, resulting in 6 fT/Hz½ sensitivity.
The atom-tuning consists of three separate feedback loops – one to control the Larmor frequency and two more to eliminate the field components (x/y) orthogonal to the pump (z). 1) The Larmor loop uses two off-resonant tones applied symmetrically, +/- 20 kHz, about the target frequency of 0.2 MHz, corresponding to a 0.3 G bias field. The differential atomic response scales with the Larmor frequency and serves as the error signal. 2) An offset in the Faraday rotation arises when the static bias field—and thus the net atomic magnetization—has a component along the probe beam (x) and is used to suppress Bx. 3) With small periodic tipping of the bias field in the second transverse direction (y), both first and second harmonics appear in the Larmor error signal. The first harmonic is proportional to static By and is used to suppress it.
Using these feedback mechanisms, along with PID control, the magnetometer is locked to the target B, and remains locked even with environmental field variations over 0.5 G. Comparison with theoretical predictions is obtained by applying periodic field variations as a function of frequency f, strength dB, and direction. The magnetometer remains well locked with f*dB within a couple Hz*G, enabling correction of large, but slow, field drifts as well as smaller 60 Hz fields, thus supporting a self-correcting OPAM for unshielded environments.
The atom-tuning consists of three separate feedback loops – one to control the Larmor frequency and two more to eliminate the field components (x/y) orthogonal to the pump (z). 1) The Larmor loop uses two off-resonant tones applied symmetrically, +/- 20 kHz, about the target frequency of 0.2 MHz, corresponding to a 0.3 G bias field. The differential atomic response scales with the Larmor frequency and serves as the error signal. 2) An offset in the Faraday rotation arises when the static bias field—and thus the net atomic magnetization—has a component along the probe beam (x) and is used to suppress Bx. 3) With small periodic tipping of the bias field in the second transverse direction (y), both first and second harmonics appear in the Larmor error signal. The first harmonic is proportional to static By and is used to suppress it.
Using these feedback mechanisms, along with PID control, the magnetometer is locked to the target B, and remains locked even with environmental field variations over 0.5 G. Comparison with theoretical predictions is obtained by applying periodic field variations as a function of frequency f, strength dB, and direction. The magnetometer remains well locked with f*dB within a couple Hz*G, enabling correction of large, but slow, field drifts as well as smaller 60 Hz fields, thus supporting a self-correcting OPAM for unshielded environments.
*DARPA Contract No. HR001124C0437 & Office of the Undersecretary of Defense, Director of Defense Research & Engineering for Modernization Quantum Science Office Contract 47QFLA23C0002.
–
Presenters
-
Sofia Kampouridou
- George Mason University