The RISTRA OPO is well suited to the generation of high energy, nanosecond pulses with good beam quality. It is also a good choice for use in non-laboratory environments because it is immune to cavity misalignment. However, given its comparatively high cost of its cavity optics, it may not be the most cost effective choice for your application. Use the criteria below to decide if the RISTRA your best choice.

• Best beam quality: If your application requires the best possible beam quality for pulse energies ≥ 2 mJ the RISTRA OPO is a good choice. For even higher energies – especially if the cavity Fresnel number (FN) is large – the RISTRA may be the only suitable solution.
• Although for FN ≈ 1, differences in beam quality for various cavity designs ideally vanish, ring cavities such as the RISTRA are still preferred because it is easier to eliminate accidental multiple resonances by taking advantage of the higher number of mirror reflections. In the RISTRA one can also use the s- and p-polarizations to optimize mirror design.

• Use in non-laboratory environments: In a planar ring or a linear cavity, slight tilts of the cavity mirrors from alignment eliminate the cavity axis. For large FN this is catastrophic for beam quality. In nonplanar rings such as the RISTRA, small tilts of a mirror merely shift the cavity axis by a small amount, and the beam quality is not strongly affected. This plus the monolithic construction of the RISTRA cavity make it insensitive to vibration. The standard spring-loaded mirror retaining rings offer convenience for switching cavity mirrors, and their three point mount minimizes mirror distortion. However, in extremely high vibration environments the springs might allow unwanted movement, in which case the mirrors may be glued in place. In either case, no mirror adjustments are necessary, making the RISTRA an excellent choice for rugged or unattended applications.
• If remote angle tuning of the crystals is desired, the RISTRA is compatible with pico motors.
• The RISTRA is compact because the crystal rotation stages and mirror mounts are integral to the RISTRA body. The dimensions of the base are 70x95 mm (2.75x3.75 in.) and the height is 115 mm (4.5 in.).
• A PZT mount for one of the mirrors is available for locking the cavity to a seed laser.
• High energy with good beam quality: Even with a RISTRA it is not always possible to achieve near-diffraction-limited beam quality. The pump beam's spatial profile, the polarization of the mixing waves, the walkoff displacement, pump pulse duration, and bandwidth of the oscillating light and pump light all affect beam quality of the RISTRA OPO. Nonetheless, compared with a conventional linear cavity or planar ring cavity the RISTRA will almost always produce a better beam.
• Large wavelength tuning range: The standard for extremely broad tuning range in nanosecond OPO’s is probably type-II BBO pumped at 355 nm. Phase matching is continuous from about 400 nm to degeneracy at 710 nm. It’s possible to tune through degeneracy but this is not commonly done. Although the RISTRA might support the > 300 nm tuning range of type-II BBO, such broadly tunable mirror coatings might be difficult to make, resulting in a low damage threshold. In addition the intra-cavity λ/2 retarders will be a more costly multi-plate design and air-gapped to tolerate high peak power. If this type of tuning range is mandatory a prism-based ring cavity is an alternative, but beam quality will typically be lower.

Here are some technical details to consider in your decision concerning use of a RISTRA for a large wavelength tuning range:

1. Crystal angular tuning range: The maximum external angular tuning range for a 10 x 10 x 15 mm crystal in the RISTRA cavity is ±10°. If you must exceed this angular range you’ll need to do something different. Using shorter crystals with smaller apertures increases the angular tuning range. SNLO QMIX can calculate phase matching parameters to help determine the required angular tuning range.
2. Typical mirror reflection bandwidths: Highly- and partially-reflecting mirror coatings for 532 nm pumped RISTRA cavity mirrors, where the angle of incidence is 32.8° for all four mirrors, can have usable reflection bandwidths > 80 nm. This is true for the mirror set used in one of our publications (A. V. Smith and D. J. Armstrong, “Nanosecond optical parametric oscillator with 90° image rotation: Design and performance,” J. Opt. Soc. Am. B 19, 1801–1814 (2002).) We’ve found the reflection bandwidths may be even greater for 1064 nm pumping, so typical RISTRA mirror sets suffice for broad tuning in some applications.
3. Multi- versus zero-order intra-cavity waveplates: Broad wavelength tuning will likely require zero-order half-waveplates. The retardation error for a multi-order waveplate, compared to an otherwise equivalent zero-order waveplate, can significantly limit tuning range.
4. Single frequency oscillation: Single-frequency injection-seeded oscillation is generally incompatible with large, continuous, tuning ranges. Even if a broadly tunable seed laser is available, the large angular tuning range required for most commonly used crystals may require re-alignment of the seed laser. In addition, the PZT assembly for the RISTRA cavity offers approximately ±7 um of mirror translation. Exceeding the PZT displacement limit necessarily involves the inconvenience of repeatedly opening the servo loop and re-acquiring lock.
5. A working example of broadband tuning: We demonstrated a 355 nm pumped type-II BBO RISTRA that tunes approximately 90 nm, from about 570 nm to 650 nm. Unfortunately we have not published results for this device. Lacking other examples, this might set a reasonable bound for a large tuning range for a commonly used, highly birefringent crystal. The cavity contained one 9 x 9 x 17 mm type-II BBO crystal and used stock zero-order waveplates centered at 626 nm. We found multi-order waveplates centered at 633 nm reduced the tuning range to about 633 ± 5 nm when the OPO was pumped 2–3 x threshold.