Schematic diagram of the fiber-optic interferometer showing the tunable IR laser, ISO, evanescent wave coupler, single InGaAs detector, and the interferometer cavity.
Experimental interferometer fringe data taken for both a sweep of cavity length at fixed wavelength (a) and a sweep of wavelength at fixed cavity length (b). In both cases, the inset magnifies the data near . In (a), the cavity length is swept by linearly ramping a voltage to the piezoelectric actuator on which the fiber endface is mounted. Small, reproducible oscillations are observed when wavelength is swept.
Power spectra data observed at the output of the photodetector amplifier, using the test laser source with and without coherence control, normalized by the coherence-control-on data between 10 and 100 Hz. For signals below 1 kHz, coherence control greatly improves performance, but artifacts related to the coherence control cause the two curves to cross above 1 kHz, and a low-pass filter must be used to remove the modulation components, which, for our laser, begin at 3.9 kHz.
Power spectra data, normalized as in Fig. 3, for the low-SSE-noise and high-power laser outputs. For both traces, the coherence control was on. Use of the low-SSE noise output improves the low-frequency performance.
Power spectra data, normalized as in Fig. 3, comparing the noise from a DFB laser with the test laser. Below 50 kHz, the test laser was clearly quieter in either configuration of the coherence control (see Fig. 3 for test laser data above 1 kHz with coherence control off).
Time series noise data sampled at a 2 kHz rate by a digital volt meter used as a digitizer. The noise at the detector is Gaussian in nature and has a standard deviation of 2.2 pm. A −24 dB/octave, 1 kHz analog low-pass filter was used as an antialiasing filter.
PSD of interferometer noise data, in absolute units of , calculated from data like those shown in Fig. 6, but for 25 s (blue curve). Also shown is the PSD of the detector dark current noise (green curve) sampled under the same conditions but with the laser turned off. The shot noise limit, corresponding to the photodetector transimpedance amplifier gain of and a quadrature detector output level of 5 V (for laser power), is shown for reference (red dashed line).
Measurement noise data in picometers, obtained as in Fig. 6, but for cavity spacing ranging from to over . The interferometer noise is constant over changes in cavity length out to a gap. The ability to operate at larger gaps with no noise penalty is desirable.
Article metrics loading...
Full text loading...