An interpreter gets distracted by every rustle in the hall, and a one-hour meeting stretches into two. The same happens with a capacitive gyroscope: to hear the nanometer-scale displacement of its proof mass, it needs perfectly tuned gaps between the plates. The smaller the sensor, the harder it is to maintain that precision. Engineers have long been looking for another way to listen to motion.

In the new work, the interpreter's role is taken over by a magnetic tunnel junction (MTJ): a sandwich of two magnetic layers with an insulator between them. When a radio-frequency current passes through it, the junction rectifies it into a DC voltage, and that voltage changes when a magnetic proof mass moves nearby.

The model, built on the resonator's equations of motion and magnetization dynamics, is calibrated against experimental data: real MTJs measuring 300×750 nm showed a sensitivity of about 20 μV/mT. Mechanical parameters are typical for MEMS resonators: mass 1 μg, resonant frequency 250 kHz, quality factor 157. In steady state, the differential voltage grows linearly with angular velocity, yielding a sensitivity of 0.001 μV/(°/s).

The estimated angular random walk (ARW) was 1800 °/√h, which corresponds to consumer-grade accuracy. For comparison, the thermal limit for this resonator gives an ARW of about 0.04 °/√h, meaning electronic readout noise dominates. It is like listening to a whisper in a noisy room: the signal is there, but the crackle of the equipment drowns it out.

In a dynamic scenario, a wind gust with a 40 kHz modulation, homodyne demodulation recovered the angular velocity with an error of about 0.05%. And if the passive proof mass is replaced with an active spin-torque nano-oscillator (STNO), the sensitivity increases by a factor of 70, while the drive amplitude drops to 150 nm, roughly 500 times thinner than a human hair.

The interpreter finally stops being distracted by rustles. Now it listens to the magnetic echo of motion, and that echo can be amplified without touching the hall itself. All that remains is to make its voice clearer.