Light intensity-voltage correlations and leakage-current excess noise in a single-mode semiconductor laser

Semiconductor lasers are particularly well suited for the implementation of pump-noise suppression, yielding a reduction of the intensity noise in the laser. In this simple picture, the maximal amount of squeezing is equal to the quantum efficiency. However, experimental results on intensity noise reduction by pump-noise suppression are usually above this limit. This discrepancy suggests that additional noise sources must be involved. Here we successfuly interpret the full noise behavior of a single-mode laser diode far above threshold by considering two excess noise sources: the leakage current fluctuations across the laser and the Petermann excess noise. We have estimated the contribution of each noise source using the results of the correlations between the laser output intensity noise and the voltage fluctuations across the laser diode (light-voltage correlations) and obtained good agreement between our theory and experimental results.

Figure 1

Fermi levels into the laser structure.

Figure 2

(Color online) Electronic equivalent circuit for the laser.

Figure 3

(Color online) Experimental setup to measure the intensity noise, the voltage noise, and the correlations between both.

Figure 4

(Color online) Experimental setup to measure the Fano factor before the spectrometer and the longitudinal side mode rejection rate. To knock from one measure to the other, we used a movable mirror.

Figure 5

Fano factor of the laser diode beam for a driving current of $80\mathrm{mA}$.

Figure 6

Experimental (big crosses) and theoretical (curves) results for Fano factor $S$. Curve 1 corresponds to the “standard” model; curves 2–4 are for the “modified” model. Curve 2 is without excess noise, 3 corresponds to the Petermann noise, 5 is for the leakage current and the Petermann noise, and 4 is without Petermann noise $(K=1)$, but with the leakage current noise.

Figure 7

Experimental (big crosses) and theoretical (curves) results for the normalized current fluctuations $J_n$ across the laser. Curve 1 corresponds to the “standard” model; curves 2–4 are for the “modified” model. Curve 2 is without excess noise, 3 corresponds to the Petermann noise, 4 is for the leakage current and the Petermann noise, and 5 is without Petermann noise $(K=1)$, but with the leakage current noise.

Figure 8

Experimental (big crosses) and theoretical (curves) results for the light-voltage correlations $C$. Curve 1 corresponds to the “standard” model; curves 2–4 are for the “modified” model. Curve 2 is without excess noise, 3 corresponds to the Petermann noise, 5 is for the leakage current and the Petermann noise, and 4 is without Petermann noise $(K=1)$, but with the leakage current noise.

Figure 9

(Color online) Experimental setup to measure the laser diode resistance value $R_d(J,F)$.

Figure 10

Laser diode resistance value versus frequency at a fixed driving current of $55\mathrm{mA}$.

Figure 11

Average laser diode resistance value between 1 and $10\mathrm{MHz}$ versus the driving current $J$.

Figure 12

(Color online) Experimental setup to measure the noise of our power supply.

Figure 13

Power-supply current noise normalized to the standard quantum limit versus the driving current $\left(J\right)$ at a fixed frequency of $6\mathrm{MHz}$.