Scattered light reduction in Sagnac Speed Meters with Tunable Coherence

This paper experimentally demonstrates that the "Tunable Coherence" technique, which controllably breaks a laser's long coherence length, effectively suppresses scattered light noise by 24.2 dB in Sagnac interferometers and offers a fundamental solution for enhancing the sensitivity of ring resonators.

The Gist

The Big Picture: Listening to the Universe's Whispers

Imagine you are trying to hear a tiny, delicate whisper (a gravitational wave from a black hole collision) in a room that is absolutely filled with a chaotic, loud party. The "whisper" is the signal scientists want to detect. The "party noise" is scattered light.

In giant laser detectors (like LIGO), a laser beam travels through a long tunnel. Sometimes, a tiny bit of that laser light hits a speck of dust or a rough spot on a mirror and bounces off in the wrong direction. It wanders around, hits something else, and eventually bounces back into the main beam. Because it took a "detour," it arrives at the wrong time, creating a confusing noise that drowns out the cosmic whisper.

For a long time, scientists have tried to fix this by building better walls (baffles) or using software to clean up the noise later. But for the next generation of super-sensitive detectors, even a single stray photon (particle of light) is too much noise.

The New Idea: "Tunable Coherence" (The Shuffling Deck)

The authors of this paper tested a clever new trick called Tunable Coherence.

The Analogy: The Perfectly Synchronized Swimmers
Imagine two swimmers in a pool. If they are perfectly synchronized, they move in perfect unison. If you try to mix them with a third swimmer who is out of sync, the water gets messy and chaotic.

In a normal laser, the light waves are like those perfect swimmers. They are all marching in step. If a stray photon bounces off a mirror and comes back, it's still marching in step with the main beam, so it interferes and creates noise.

The Solution: The Random Shuffle
The scientists decided to break the synchronization on purpose. They used a device to "jitter" the laser light using a Pseudo-Random Noise (PRN) sequence. Think of this as a complex, random code stamped onto the laser light, like a unique barcode that changes a billion times a second.

• The Main Beam: Carries the code.
• The Stray Light: Also carries the code, but because it took a detour, the code is delayed.

When the main beam and the stray light meet again, their "barcodes" don't match. It's like trying to fit a puzzle piece from the left side of a box into the right side of a different box. They don't click together. Because they don't match, they don't interfere, and the noise disappears.

The Experiment: The Sagnac Speed Meter

The scientists tested this on a specific type of detector called a Sagnac Speed Meter.

• The Setup: Imagine a race track where two runners (light beams) run in opposite directions around a loop. They start together, run the loop, and meet again.
• The Problem: In these loops, light can easily bounce from one runner to the other (backscatter), causing them to trip over each other.
• The Test: They built a tabletop version of this race track. They intentionally created a "stray light" problem by reflecting a tiny bit of the beam back into the main path.

The Results:
They turned on their "jittering" code.

• Without the code: The stray light made a huge mess (noise).
• With the code: The stray light was suppressed by 24.2 decibels.
  • Analogy: This is like turning a roaring jet engine down to the sound of a quiet conversation. It's a massive reduction in noise.

Why This Matters for the Future

The paper shows that this trick works not just for simple straight-line detectors, but also for the complex "loop" detectors (Sagnac) that future observatories might use.

The Catch (The Limitations):
It's not magic; it has rules.

1. The "Chip" Size: The random code is made of tiny blocks called "chips." If the stray light's detour is shorter than the size of one chip, the codes still match, and the noise remains. The stray light must travel far enough to get "out of sync" with the code.
2. Perfect Timing: You have to be very careful to match the timing of the main beam and the "local" reference beam. If they are slightly off, you might accidentally filter out the signal you want to hear, not just the noise.

The Bottom Line

This paper proves that by intentionally making a laser light "jitter" with a random code, we can trick stray light into thinking it doesn't belong. It's like giving every particle of light a unique ID card; if the ID card is expired or delayed, the particle is ignored.

This is a huge step forward for building the next generation of gravitational wave detectors, allowing them to hear the faintest whispers of the universe without being drowned out by the noise of their own equipment.

Technical Summary

1. Problem Statement

Laser interferometers, particularly those used in gravitational wave detection (e.g., LIGO, Virgo) and ring resonators (e.g., gyroscopes), are limited in sensitivity by scattered light noise.

• Mechanism: Light leaves the main beam, scatters off optical surfaces, travels an unintended path, and re-couples into the readout. This introduces non-linear phase noise and power imbalances.
• Specific Challenge in Sagnac Topologies: While Sagnac Speed Meters offer advantages for quantum noise reduction (as speed is a quantum non-demolition observable), they suffer from a unique issue called backscatter. In ring resonators and Sagnac interferometers, light from one propagation direction can scatter directly into the counter-propagating beam. This creates phase noise and power offsets that are difficult to mitigate with current post-processing techniques.
• Limitation of Current Methods: Existing solutions (baffles, active controls, post-processing) are insufficient for the extreme sensitivity requirements of next-generation observatories (e.g., Einstein Telescope, Cosmic Explorer), where even a single scattered photon can be problematic.

2. Methodology: Tunable Coherence

The authors experimentally demonstrate Tunable Coherence, a technique that fundamentally alters the laser's coherence properties to suppress interference from scattered light.

• Core Concept: The laser's long coherence length is broken in a controlled manner using high-speed phase modulation driven by a Pseudo-Random Noise (PRN) sequence (specifically m-sequences).
• Mechanism of Suppression:
  • The PRN modulation creates a "pseudo-white-light" interferometer.
  • Interference only occurs if the relative path delay between two beams is less than the length of a single "chip" in the PRN sequence ($d_{chip} = c/f_{PRN}$).
  • If scattered light travels a path difference greater than one chip length, the modulation sequences become uncorrelated (low auto-correlation), and the interference is suppressed.
  • The theoretical suppression limit is $1/n_{chips}$, where $n_{chips}$ is the sequence length.
• Experimental Setup:
  • Topology: A tabletop Sagnac interferometer (circumference ~343 cm).
  • Modulation: An Electro-Optical Modulator (EOM) applies a $\pi$-depth phase modulation at $f_{PRN} = 1$ GHz using m-sequences ranging from 7 to 2047 chips.
  • Scattered Light Simulation: A low-reflectivity mirror ($R \approx 0.2$) coupled light from the counter-clockwise beam into the clockwise beam, creating a controlled scattered light tone with a variable delay ($\tau_{sc}$) via an optical delay line.
  • Readout: A Balanced Homodyne Detector (BHD) was used. Unlike Michelson setups where arms are matched, the Sagnac setup required the Local Oscillator (LO) to be delay-matched to the signal beam exiting the interferometer. The LO was derived from the input beam before the interferometer to avoid self-interference.

3. Key Contributions

• First Demonstration in Sagnac Topology: This is the first experimental proof that Tunable Coherence is viable for Sagnac interferometers, extending its applicability beyond Michelson topologies.
• Analytical Framework for Backscatter: The authors derived analytical conditions for suppressing backscatter in ring resonators. They established that backscatter can be suppressed if the relative delay between counter-propagating beams at scattering surfaces does not match an integer multiple of the sequence's recoherence length ($d_{coh} = n_{chips} \cdot d_{chip}$).
• Quantitative Suppression: The experiment achieved a scattered light suppression of 24.2 dB using a 1023-chip sequence.

4. Results

• Delay Dependence: Suppression increases as the relative delay of the scattered light exceeds the chip length ($d_{chip} \approx 30$ cm for 1 GHz). Suppression plateaus once the delay is sufficient to fully decorrelate the sequences.
• Sequence Length Dependence: Suppression scales with sequence length (theoretically $20 \log_{10}(n_{chips})$). The experiment showed increasing suppression up to the 1023-chip sequence, reaching the 24.2 dB limit.
• Signal Integrity: A minor reduction (2–3 dB) in the simulated signal amplitude was observed, attributed to a slight delay mismatch between the LO and the signal beam due to manual alignment limitations.
• Limitations Observed:
  • Residual Noise: Full suppression was not achieved due to imperfections in the modulation depth (requiring exactly $\pi$) and potential electronic crosstalk in the high-voltage amplifiers.
  • Bandwidth Constraints: The measurement bandwidth is limited by the PRN sequence repetition rate (typically < 500 kHz in this setup), preventing the observation of higher-frequency effects.
  • Non-Rectangular Modulation: Finite rise/fall times in the modulation hardware degraded the auto-correlation function, causing suppression to saturate earlier than theoretical predictions for longer sequences.

5. Significance and Future Implications

• Fundamental Solution: Tunable Coherence offers a fundamental, hardware-based solution to scattered light and backscatter noise, rather than relying on post-processing or mechanical baffles.
• Impact on Future Detectors: This technique is highly relevant for next-generation gravitational wave observatories (Einstein Telescope, Cosmic Explorer) and high-precision ring laser gyroscopes (e.g., ROMY), where backscatter is a critical noise source.
• Design Constraints: The analysis provides specific design rules for ring resonators: mirror placement and cavity lengths must be engineered to ensure that backscattered light accumulates a delay mismatch relative to the counter-propagating beam that exceeds the chip length ($d_{chip}$).
• Trade-offs: While it effectively suppresses phase noise from backscatter, it does not resolve power imbalances caused by loss in one propagation direction. However, by eliminating the phase noise component, it significantly improves the signal-to-noise ratio for speed-meter topologies.

In conclusion, the paper validates Tunable Coherence as a powerful tool for mitigating scattered light in complex interferometric topologies, paving the way for more sensitive quantum-limited measurements in future astrophysical and geophysical instruments.