Fig. 1. Experimental setup of MIR upconversion detection based on the external-cavity pumping. An EDFL at 1550 nm and a YDFL at 1064 nm are used to provide initial light sources for subsequent nonlinear frequency conversion. The two fiber lasers, operating at the single-longitudinal mode, are first used to prepare the MIR signal at 3.4μm based on the DFG in a PPLN1 crystal. Then, the generated MIR beam is injected into an optical cavity for implementing the sum-frequency generation. The cavity comprises four mirrors and is stabilized with a digital locking unit based on a programmed FPGA. Under the locked state, the pump power can significantly be enhanced within the cavity. After passing through a series of spectral filters, the upconverted signal is coupled into an SMF before being detected by an SPCM or an MPPC. YDFA and EDFA, Yb- and Er-doped fiber amplifiers; DM, dichroic mirror; PBS, polarization beam splitter; L, lens; Atten, attenuator; SM, silver mirror; Col, collimator; ISO, isolator; HWP and QWP, half-wave and quarter-wave plates; M, cavity mirror; PZT, piezoelectric actuator; PD, photodiode; PC, computer; HV, high-voltage amplifier; PS, power sensor; NF, notch filter; LP, SP, and BP, long-, short-, and bandpass filters.

Fig. 2. (a) Photo of the enhancement optical cavity, indicating the materials of each part. (b) Physical layout of the bow-tie cavity. The distances D1 and D2 are two critical parameters for the cavity design. (c) The radius of the beam waist within the nonlinear crystal as a function of D1 and D2. The black point denotes the values used in the experiment, corresponding to a beam radius of 69μm. Note that the area in white represents the unstable region for an optical cavity. (d) Evolution of the beam size as the light propagates along the cavity, with and without the presence of the crystal inside the cavity. The origin is defined at the center of the crystal. The shaded area indicates the occupied space by the 25-mm-length nonlinear crystal.

Fig. 3. (a) Recorded traces of the reflection (PD1) and transmission (PD2) lights from the optical cavity while scanning the PZT at a rate of 20 Hz. The peaks correspond to the resonating points for the fundamental spatial mode. (b) Enlarged illustration for the peak, along with the error signal for the dither locking. (c) Measured intensity for the cavity transmission during the stabilization process. The locking is engaged at 18 s, as indicated by the red triangle. The presented disturbing drops are ascribed to the intentional perturbation for testing the re-locking feature of the digital feedback unit.

Fig. 4. (a) SFG power as a function of crystal temperature. (b) Captured beam profile for the SFG light. (c) Intrinsic conversion efficiency and total detection efficiency increase as augmenting the intracavity power. (d) Measured background count and NEP as a function of the intracavity power.

Fig. 5. Measured photon counts varying with the incident power is increased in two cases of using optical detectors based on SPCM and MPPC, which indicate detection dynamic ranges of 15 and 30 dB in the linear response regime, respectively. The solid lines are fitted to guide the eyes.