1. Introduction

Q-switched single-longitudinal-mode (SLM) lasers in the nanosecond regime with a stable repetition rate are widely used in a master oscillator and power amplifier, coherent laser radar, ranging, and for nonlinear applications. Laser-diode- (LD-) pumped Cr⁴⁺:YAG passively Q-switched Nd³⁺:YAG lasers can supply SLM operation, but the timing jitter is excessively large for the above applications [1,2,3]. Research has been undertaken to reduce timing jitter by modulation of the pump light to passively Q-switched (PQ) lasers [4,5], but the jitter is still excessive for many applications. Combined actively and passively Q-switched (CAPQ) lasers can provide SLM pulses with minimal timing jitter. An earlier version of CAPQ lasers pumped by a helical xenon flashlamp [6] and stimulated by a discharge tube [7] was used to control the pulse width. A Nd:YAG laser was reported in 1978 [8] that simultaneously uses a Pockels cell and flowing dye is a CAPQ laser pumped by a flashlamp, and operates in a SLM with a low timing jitter. The longitudinal and transverse mode selection for this CAPQ laser was achieved by use of a resonant reflector and an aperture. In another flashlamp-pumped CAPQ laser, the single-mode selection was achieved with a Fabry–Perot (FP) etalon and the twisted mode method [9].

Recently, by use of a LiNbO₃ Pockels cell and a Cr⁴⁺:YAG crystal, the timing jitter was compressed to 65 ps in a LD-pumped CAPQ Nd³⁺:YVO₄ microchip laser and the average output power was 9 mW at a pump power of 350 mW, corresponding to a 2.6% conversion efficiency [10,11]. In another experiment a LD-pumped CAPQ laser with a ring cavity was used for SLM operation [12]. We present a simple configuration of a LD-pumped watt-level SLM CAPQ laser with a wavelength of 1064 nm that simultaneously uses a codoped Cr⁴⁺,Nd³⁺:YAG crystal as the gain medium and mode selector. The experimental results show that SLM pulses have lower timing jitter, a narrower pulse width, a high repetition rate, stable pulse-to-pulse amplitude, and high-power conversion efficiency. We report the results of a comparison of actively Q-switched (AQ) Nd:YAG, PQ, and CAPQ Cr,Nd:YAG lasers with the same configuration.

2. Experimental Setup

The approach used to achieve SLM selection in the optical cavity and a stable pulse repetition rate is based on use of the loss grating effect [1] of a codoped Cr,Nd:YAG crystal, active Q-switch control, and precise temperature control. A schematic of the CAPQ laser system is shown in Fig. 1. Two kinds of Q switches are incorporated into the laser design: a passive absorber Q switch and a standard acousto-optic modulator (AOM). The purpose of simultaneous use of the AOM active Q switch is to reduce the timing jitter of the passive Q switch from several hundred microseconds to a few hundred nanoseconds. Because of the nonlinear saturation characteristics of the Cr,Nd:YAG laser, the passive Q switch saturable absorber also acts as a cavity longitudinal mode selector [13,14].

The details of the optical design are as follows. The plane–plane laser cavity consisted of a rear mirror that was coated for high transmission at an 808 nm diode pump wavelength and total reflection at a lasing wavelength of 1064 nm, and an output coupler coated for transmission of 30% at 1064 nm. Because of the large dimensions of the AOM, the cavity length was minimized to 66 mm to maximize longitudinal mode spacing. The experiments below were conducted on a Cr,Nd:YAG crystal that was grown along the [111] crystallographic axis and each facet was coated for high transmission at 1064 nm. The 1.0 at.% Nd³⁺ and 0.1 at.% Cr⁴⁺-codoped crystal with dimensions of 4.9 mm × 4.9 mm × 5.6 mm having small-signal transmissions of 75.5% at 1064 nm and 49.1% at 808 nm was inserted inside the laser resonator near the rear mirror. The codoped crystal was mounted in a water-cooled copper block, and its temperature was maintained at 17°C. In another comparison of an actively Q-switched laser with the same configuration, the Cr,Nd:YAG laser was replaced by a Φ₂ mm × 5 mm Nd:YAG crystal and each facet was coated for high transmission at 1064 nm. The Nd:YAG crystal transmission at 1064 nm was 99.6% and at 808 nm it was 3.3%. A water-cooled AOM (QSGSU-5 from No. 26 Research Institute, Chongqing, China) with low insertion loss, driven at a rf of 27 MHz, a maximum rf power of 50 W, and a 4.8–16.2 kHz modulation frequency range was used as the active Q switch and placed within the cavity. The modulator was 53.3 mm in length, and the rod, the modulator, and the output coupler were spaced approximately 1 mm apart. The pump laser used was a 30 W fiber-coupled cw LD array (CGLA-B-30-05, Changchun Institute of Optics, Fine Mechanics and Physics, Changchun, China) operated at 807.3 nm and its temperature was maintained at 25°C. The optical fiber diameter was 400 μm with a numerical aperture of 0.22. The pump beam that emitted from the fiber pigtail was collimated and focused into the laser rod by means of a rear mirror with a pair of 1:1 plano–convex focusing lenses with a 30 mm focal length. Thus, the largest dimension of the spot size produced by the pump on the input face of the Cr,Nd:YAG rod was approximately 400 μm, enabling all the light to be pumped into the TEM₀₀ lasing-mode spot size of 425 μm at a maximum of 20 W pump power. In addition, the dynamic aperture effect of the Cr,Nd:YAG laser also prevents high-order transverse modes from oscillating. Therefore, TEM₀₀ operation can be ensured.

 

Fig. 1. CAPQ Cr,Nd:YAG laser setup.

Download Full Size | PDF

3. Results

To achieve stable repetition-rate, single-frequency operation of the device, we utilized a CAPQ excitation scheme.

It has been known for quite some time that passive Q switches tend to act as mode selectors [13,14]. Longitudinal-mode selection in a laser takes place while the pulse builds up from noise. There are two important parameters that determine the spectral output of a laser: one is difference in gain or loss between axial modes, and another is the number of round trips it takes for the pulse to build up from noise. With a 90 mm optical cavity length and a power ratio of 2 between pump power density and saturation power density, one can calculate that the smallest absorption coefficient difference between the zero-order axial mode and the nearest adjacent axial mode is 0.0077 cm^-1 [1]. Our experiments showed that single frequency can be obtained without an active Q switch even at a maximum pump power, so it is reasonable to suppose that the laser pulse build-up time is no less than the actively Q-switched pulse build-up time of 150 ns. In this case, for a 90 mm optical length cavity the time delay amounts to approximately 250 round trips for the energy to build up. This means that the dominant mode should have at least ten times greater peak power than any other mode.

In comparison with an AQ laser, the threshold of the PQ laser is not a fixed value, but rather a dynamic range, since the process is statistical [10]. In a CAPQ laser, if the inversion passes this range rapidly, the uncertainty in time when the pulse is emitted will be minimal. The amount of time necessary for removal of the inserted losses in the CAPQ laser is determined by the drive electronics and passive absorber. The jitter will therefore be determined not only by the jitter in the drive electronics and the build-up time of the laser pulse but also by the passive Q-switch repetition period. The optimum repetition period of the CAPQ laser is approximately 1.5 times as long as the natural repetition period of the PQ laser.

Since the build-up time in an AQ laser is in the range of a few microseconds, it should be possible to obtain jitter in the range of a few hundred nanoseconds because of the statistical variations at the start of the build-up time.

The output Q-switched pulse was monitored simultaneously by both an oscilloscope (TDS 3052, Tektronix, Richardson, TX, USA) with a 1 ns rise time Si p-i-n photodiode (DET200, Thorlabs, Newton, NJ, USA) and a FP interferometer (иT51-30, OMO, USSR). The FP had 60 MHz resolution at 1064 nm, 10 GHz free spectral range, and showed the spectrum of the Q-switched pulse on a computer screen, which was detected by an IR-sensitive CCD camera. Figure 2, which was taken by a digital camera (Canon PowerShot S50) with a shutter speed of 1/8 s shows the output spectrum resolved by the FP when the CAPQ method was applied. The photograph in Fig. 2 shows spectra with a linewidth of 530 MHz. At a 20 W maximum pump power, corresponding to an 8.6 W absorbed pump power, the CAPQ method gave a 1036 mW time-averaged power output (146 μJ energy) and a 530 MHz bandwidth for single axial mode operation. Figure 3 shows a typical CAPQ pulse shape with an output energy of 146 μJ and a FWHM duration of 8.31 ns. The corresponding peak power is 17.6 kW.

 

Fig. 2. Output SLM spectra of the CAPQ Cr,Nd:YAG laser resolved by the FP with a linewidth of 530 MHz.

Download Full Size | PDF

 

Fig. 3. Oscilloscope 512 average pulse traces of the CAPQ Cr,Nd:YAG laser with a FWHM duration of 8.31 ns, a repetition rate of 7.10 kHz, and an absorbed pump power of 8.6 W.

Download Full Size | PDF

A drawback of PQ lasers is a rather large timing jitter of the pulse train that arises from both the pumping instabilities and the noise of the spontaneous emission in the active medium. A histogram is a practical method to use for measurement of the timing jitter of pulses while the waveform persistence of oscilloscope is turned to infinite. This technique not only shows worst-case jitter, but also gives a perspective for jitter distribution. The experimental results shown in Fig. 4 demonstrate that the CAPQ method reduced the timing jitter significantly in comparison with just passive Q switching.

Figure 5 shows the average output power of an AQ laser and a PQ laser with the same configuration. Although the threshold of the AQ Nd:YAG laser is low compared with the PQ laser, the SLM output can be obtained only near the threshold. The corresponding single frequency conversion efficiency was 1.6%. In the PQ laser, the SLM can be obtained from threshold to maximum pump power. The conversion efficiency of the PQ laser at the maximum pump power is 5.4%. The PQ laser is not optimized and the Cr,Nd:YAG crystal is, which is currently available from Shanghai Institute of Optics and Fine Mechanics. The threshold of the CAPQ laser is high because of the codoped crystal with a low intensity transmission of 75.5% at a wavelength of 1064 nm and with a high transmission of 49.1% at a wavelength of 808 nm.

 

Fig. 4. (a) PQ laser operating at approximately 12 kHz with infinite persistence turned on for 2 min. (b) Typical single-pulse train for the CAPQ laser operating at 7.10 kHz. Both lasers correspond to 8.6 W of absorbed pump power.

Download Full Size | PDF

 

Fig. 5. Average output power of the AQ laser and the PQ laser with the same configuration.

Download Full Size | PDF

 

Fig. 6. Average output power of the CAPQ laser at different AOM repetition rates.

Download Full Size | PDF

Figure 6 shows the average output power of the CAPQ laser at different AOM repetition rates. It is obvious that the maximum output with SLM operation can be obtained at 7 and 13 kHz AOM repetition rates pumped at a 20 W level. The AOM repetition rate is in integral multiples of the CAPQ laser repetition rate that is determined by the natural repetition rate of a PQ laser. The two power peaks approximately correspond to the same CAPQ laser repetition rate at 6.9 kHz. The AOM does not influence the SLM operation, and the 1064 nm laser output power consumption of the AOM was less than 50 mW. The conversion efficiency of the CAPQ laser at the maximum pump power is 5.2%.

A comparison of AQ Nd:YAG, PQ Cr,Nd:YAG, and CAPQ Cr,Nd:YAG lasers with the same configuration at the threshold pump power and maximum output power is listed in Table 1. For the AQ Nd:YAG laser, single-frequency output is possible only at the threshold pumping level with careful adjustment of the pump power, and the pulse-to-pulse amplitude jitter is 100%. For higher pump power, the AQ laser has a multifrequency output. For the PQ Cr,Nd:YAG laser, stable single-frequency output can be obtained even at a maximum pump power. However, the timing jitter is large, 268 μs at the threshold pump power and 37.6 μs at the maximum output power. In contrast, with the CAPQ stabilization scheme, reliable single-frequency operation was also obtained, amplitude fluctuation remained below 4%, timing jitter was approximately 0.40 μs, and the pulse width was considerably shortened to 8.31 ns. In Table 1, all the timing jitter was measured at 2 min intervals. An optimal maximum repetition rate of 7.10 kHz was obtained, which we determined by using the natural repetition rate of 12.0 kHz of a passive Q switch at 20 W pump power. The CAPQ laser repetition period is always one time larger and two times smaller than the natural repetition period of a PQ laser. When the AOM repetition rate approaches these parameters, the timing jitter increases rapidly. For the 11.4 W low pump level, the CAPQ laser repetition rate is one third of the 6.15 kHz AOM frequency. The stable single axial mode without mode hopping can be maintained for approximately 7 min at 20 W pump power.

If we took advantage of a more stable pump power, the CAPQ laser could be improved by use of a prepump technique, resulting in reduced timing jitter and mode hopping and a more stable pulse-to-pulse amplitude.

Table 1. Comparison of the AQ Nd:YAG, PQ, and CAPQ Cr,Nd:YAG Lasers

View Table

4. Conclusion

In conclusion, we have demonstrated a simple hybrid Q-switching technique that uses a combination of acousto-optic Q switch and saturable absorber, which yields reliable, stable single-frequency operation of a Q-switched laser with a repetition rate timing jitter of less than 0.40 μs for a LD-array-pumped Cr,Nd:YAG laser. This allows most of the available single-frequency energy to be extracted as AQ output.