Squeezing in the Audio Gravitational-Wave Detection Band

We demonstrate the generation of broadband continuous-wave optical squeezing from 280 Hz–100 kHz using a below-threshold optical parametric oscillator (OPO). The squeezed state phase was controlled using a noise locking technique. We show that low frequency noise sources, such as seed noise, pump noise, and detuning fluctuations, present in optical parametric amplifiers, have negligible effect on squeezing produced by a below-threshold OPO. This low frequency squeezing is ideal for improving the sensitivity of audio frequency measuring devices such as gravitational-wave detectors.

Figure 1

Schematic of the experiment. The experiment was operated in both OPO and OPA modes. The OPA seed power was varied using a variable attenuator (VA). The OPO cavity was isolated from backscatter off the photodetectors using a Faraday isolator (FI). The control electronics for the homodyne detection phase are indicated by dashed lines. SA, spectrum analyzer; BPF, band pass filter; ED, envelope detector; LPF, low pass filter; G gain stage; SHG, second harmonic generator; OPO/A-optical parametric oscillator/amplifier; PZT, piezoelectric transducer; PBS, polarizing beamsplitter; $\lambda /2$-half-wave plate; DC, dichroic mirror; PD, photodetector.

Figure 2

Measured noise spectra for ($a$) the quantum noise limit, ($b$) the squeezed light, and ($c$) the electronic noise of the homodyne detection system. The traces are pieced together from three FFT frequency windows: 100 Hz–3.2 kHz, 1.6 kHz–12.8 kHz, and 3.8 kHz–100 kHz. Each point is the averaged rms value of 500, 1000, and 2000 measurements made in the respective ranges. The RBW of the three windows was 8, 32, and 128 Hz, respectively. The electronic noise was $-12\mathrm{d}\mathrm{B}$ below the quantum noise from 10–100 kHz. The 20 kHz peak arises from the homodyne locking signal. Peaks at 50 Hz harmonics are due to electrical mains supply.

Figure 3

The squeezed state at 11.2 kHz as the phase of the homodyne is varied. $\mathrm{R}\mathrm{B}\mathrm{W}=1\mathrm{k}\mathrm{H}\mathrm{z}$, $\mathrm{V}\mathrm{B}\mathrm{W}=30\mathrm{H}\mathrm{z}$. Electronic noise (9 dB below SNL) was subtracted from the data.

Figure 4

The OPO spectrum 100 Hz–3.2 kHz without (a) and with (b) the Faraday isolator between the OPO cavity and homodyne detector. $\mathrm{R}\mathrm{B}\mathrm{W}=8\mathrm{H}\mathrm{z}$; number of rms averages for (a) 400 and for (b) and (c) 500. Electronic noise (not shown) was not subtracted.

Figure 5

The OPA spectrum 2–100 kHz with different OPA seed powers. The variance of the seed beam with power $6\mu \mathrm{W}$ was less than 4 dB above the SNL across the spectrum. $\mathrm{R}\mathrm{B}\mathrm{W}=128\mathrm{H}\mathrm{z}$. Number of rms averages, 1000, except for $6\mu \mathrm{W}$ seed power which had 500 rms averages. Electronic noise (at $-12\mathrm{d}\mathrm{B}$) was subtracted from all traces.

Figure 6

The average noise power from 5–6 kHz as a function of seed power, experimental data indicated by “x”, model fit given by line. Electronic noise (at $-12\mathrm{d}\mathrm{B}$) was subtracted from all data.