⚛️ quantum physics

Heterodyne position detection of an optomechanical system

This paper presents a robust heterodyne position detection scheme for optically levitated particles implemented via FPGA-based digital demodulation, which outperforms standard homodyne methods by eliminating phase-wrapping distortion, ensuring immunity to local oscillator power drifts, and maintaining stability against strong parasitic back-reflections.

Original authors: Daniel Tandeitnik, Gabriel Dias, Thiago Guerreiro

Published 2026-07-14

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Daniel Tandeitnik, Gabriel Dias, Thiago Guerreiro

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to listen to a tiny, jittery dancer (a microscopic particle) who is spinning in a beam of light. To know exactly where the dancer is, you need to measure the tiny ripples they make in the light. For a long time, scientists used a method called homodyne detection to do this. Think of it like trying to hear a whisper by holding a tuning fork next to your ear. It works, but it has three big problems that make it frustrating for a curious teenager (or a precise scientist) to use.

First, the "whisper" is often drowned out by a loud, annoying echo bouncing off the walls of the room (parasitic back-reflections). In the old method, this echo messes up the tuning fork, making the whole system lose its rhythm. Second, if the dancer moves too far, the signal gets "wrapped up" like a tangled headphone cord, distorting the shape of the movement. Third, if the light bulb flickers just a tiny bit, the measurement changes, forcing you to constantly re-tune your equipment.

In this paper, the authors propose a new way to listen: heterodyne detection combined with a super-fast digital brain (an FPGA). Instead of holding a tuning fork, they give the light a tiny, constant "hum" (a frequency shift of 10 MHz) before it hits the detector. Then, they use digital math to separate the dancer's movement from the noise.

Here is why this new method is a game-changer, based on what the authors actually measured and simulated:

1. The "Echo" Problem is Solved
In the old method, if the room was full of echoes (strong parasitic fields), the system would get confused and drift. The authors show that with their new digital method, they can filter out those echoes like a noise-canceling headphone. They measured this directly: even when the "echo" was stronger than the signal, the new method kept working perfectly. In fact, they found that the new method gave a 10 dB better signal-to-noise ratio than the old one when the light was at its strongest. That's like hearing a pin drop in a hurricane while the old method was still struggling to hear a shout.

2. No More Tangled Headphones (Phase Wrapping)
When the dancer moves a lot, the old method's signal gets squashed and distorted, like a sine wave that has been folded over itself. The authors simulated this and found that it made the signal look like it had extra, fake "harmonics" (ghost notes) that weren't really there. This made it impossible to know exactly how far the dancer moved.
The new method, however, produces a signal that is perfectly straight and linear. When they tested this by shaking the particle with an electric voltage, the old method gave a wobbly, non-linear response that got worse as they shook it harder. The new method gave a perfectly straight line, meaning they could measure the movement accurately no matter how much the particle jumped around.

3. The "Flickering Bulb" Problem is Gone
In the old setup, if the laser power drifted even slightly, the measurement scale changed. It was like trying to measure a room with a ruler that shrinks and grows depending on the temperature. The authors proved that their new method is immune to this. They changed the brightness of the local light source and watched the measurement scale. For the old method, the scale changed linearly with the brightness. For the new method, the scale stayed exactly the same, no matter how the light drifted. This means you don't have to stop your experiment every hour to re-calibrate your ruler.

How They Did It
The team didn't just guess; they built a real experiment with a tiny silica particle (156 nm wide) trapped in a vacuum. They used a special computer chip (an FPGA) to do the math in real-time, processing the data in about 600 nanoseconds. This is fast enough to not only measure the particle but also to give it a "cooling" push to stop it from jittering, a technique known as feedback cooling.

The Bottom Line
The authors demonstrate that this new heterodyne approach is robust, linear, and stable. It works even when the "echoes" are loud, it doesn't get tangled when the particle moves far, and it doesn't care if the light flickers. While they tested this on a levitated particle, the paper suggests this digital trick could be used in any system where you need to measure motion by looking at light, from tiny sensors to massive gravitational wave detectors. It turns a finicky, high-maintenance measurement into something that just works, letting scientists focus on the physics rather than fighting their equipment.

Technical Summary: Heterodyne Position Detection of an Optomechanical System

Problem Statement
Optomechanical systems, ranging from gravitational-wave interferometers to optically levitated particles, rely on the transduction of mechanical motion into optical phase shifts. The standard method for extracting this phase information is homodyne detection, where a scattered field (SF) interferes with a phase-locked local oscillator (LO). The authors identify three critical limitations in the homodyne approach, particularly in high-efficiency regimes:

Phase-Locking Instability: Strong parasitic back-reflections (PFs) from optical components can dominate the signal, causing the active phase-locking loop to drift or lock to incorrect operating points, thereby destabilizing the measurement.
Nonlinearity and Phase Wrapping: The homodyne signal is sinusoidal with respect to the phase difference ( $\sin(\Delta\phi)$ ). Large mechanical excursions lead to phase wrapping and nonlinear distortion, introducing odd-order harmonics and biasing calibration procedures based on power spectral density (PSD).
Calibration Drift: The homodyne transduction gain scales linearly with the optical amplitudes of both the LO and SF ( $A_{LO}A_{SF}$ ). Consequently, slow drifts in optical power directly alter the calibration factor, necessitating frequent recalibration and complicating long-duration experiments.

Methodology
The authors propose and experimentally demonstrate a heterodyne detection scheme combined with digital In-phase and Quadrature (IQ) demodulation implemented in real-time on a Field-Programmable Gate Array (FPGA).

Experimental Setup: The system utilizes a 156 nm silica nanoparticle optically levitated in a vacuum chamber. The setup splits the back-scattered field and a local oscillator (derived from the trapping beam) into two parallel detection branches: a standard homodyne branch and a heterodyne branch.
Heterodyne Mechanism: In the heterodyne branch, an acousto-optic modulator (AOM) shifts the LO frequency by 10 MHz ( $\Delta\omega$ ). The resulting beat signal is processed digitally.
Digital Signal Processing: The FPGA pipeline (implemented on a Red Pitaya STEMLab 125-14) performs the following steps:
1. Mixing: The signal is mixed with digital references $\cos(\Delta\omega t)$ and $-\sin(\Delta\omega t)$ to generate I and Q channels.
2. Filtering: A Cascaded Integrator-Comb (CIC) filter removes the 20 MHz image frequency. Crucially, first-order IIR high-pass filters remove quasi-DC offsets caused by parasitic fields and slow LO phase drifts.
3. Phase Extraction: The phase is recovered via $\arctan(I/Q)$ , followed by a digital phase unwrapper and a final high-pass filter to isolate the mechanical motion.
Comparison: The authors simultaneously measure the same trapped particle using both schemes to directly compare performance regarding linearity, calibration stability, and signal-to-noise ratio (SNR).

Key Contributions and Results
The paper provides experimental quantification of three specific advantages of the heterodyne scheme over homodyne detection:

Robustness Against Parasitic Fields:
- In the homodyne scheme, strong parasitic reflections prevent stable phase locking, causing the operating point to drift with the parasitic phase.
- In the heterodyne scheme, parasitic contributions appear as quasi-DC offsets in the I/Q channels. These are effectively removed by high-pass filtering, allowing the system to remain robust even when parasitic power exceeds the signal power.
- Result: The heterodyne scheme demonstrated a 10 dB improvement in SNR at maximum LO amplitude compared to homodyne, directly attributed to the suppression of parasitic field noise that contaminates the homodyne measurement.
Linear Response and Elimination of Phase Wrapping:
- Homodyne detection exhibits sinusoidal nonlinearity. When the particle displacement approaches $\lambda/4$ , the signal undergoes phase wrapping, redistributing power into odd harmonics (e.g., a strong third harmonic at 156 kHz for a 52 kHz drive).
- This nonlinearity caused a non-monotonic force-voltage response during electrode calibration, leading to significant underestimation of the calibration factor.
- Result: The heterodyne scheme produced a strictly linear displacement signal free of phase wrapping. Electrode calibration yielded a linear force-voltage relationship across the full voltage range, enabling accurate calibration even at large drive amplitudes.
Intrinsic Immunity to Power Drifts:
- The homodyne calibration factor ( $C_{VM}$ ) scales linearly with LO amplitude. Experimental data showed $C_{VM}$ varying significantly as the LO power was adjusted.
- The heterodyne phase readout ( $\arctan(I/Q)$ ) is independent of the absolute optical amplitudes of the LO and SF.
- Result: The heterodyne calibration factor remained constant despite variations in LO amplitude, eliminating the need for periodic recalibration due to optical power drifts.

Significance
The authors conclude that heterodyne detection with digital IQ demodulation offers a robust and practical alternative to homodyne detection for phase readout in optomechanical systems. The method is particularly significant for configurations where parasitic back-reflections cannot be fully suppressed or where stable, drift-resilient calibration is required for long-duration precision measurements (e.g., feedback cooling, force sensing, and impulse measurements).

The implementation is platform-agnostic; while demonstrated on an optically levitated nanoparticle, the underlying signal processing can be adapted to any optomechanical system where mechanical motion is encoded in the phase of a scattered or transmitted field. The total pipeline latency of approximately 600 ns is compatible with velocity-damping feedback cooling for mechanical frequencies up to 200 kHz, and the system allows for runtime reconfiguration of measurement range and resolution without FPGA reprogramming.

More like this