There are many performance data that specify a complex unit like our DLC202. We have concentrated on ergonomics, but always with the ultimate laser linewidth in mind. The diode injection current noise is so low that in almost all cases, the laser performance is determined by the optical characteristics of the external cavity, not the MOGLabs controller.
The curve below shows the diode injection current noise (or lack thereof). Measured at 50mA into 50 ohms (orange trace) and zero current (blue trace). The current noise level is well below -145dBm (0.25nA/√Hz) for nearly all of the spectrum (measured at 1Hz RBW).
Typical frequency noise spectrum with a CEL001 cateye laser. The first resonance with our CEL002 laser occurs at around 30kHz.
The figure below shows a self-heterodyne beatnote for a single diode laser, locked to a rubidium transition. The beatnote width was roughly 80kHz (FWHM), 35kHz (rms). Our linewidth measurements are described in Appl. Opt., 48 6961-6966 (2009).
In another experiment, we demonstrated a wide-bandwidth frequency locking technique based on electromagnetically induced transparency (EIT) using FM sideband locking. The figure below shows the beatnote frequency with two lasers locked to transitions in Rb87, with different hyperfine ground states separated by 6.8GHz. One MOG was locked to a saturated absorption peak, the other to the EIT resonance. The beatnote, which is a measure of the relative frequency stability, has a width of 4kHz for a 22.5s average, and less than 1kHz for a single sweep of 0.2s. The work was published in Appl. Phys. Lett., 90 171120 (2007).
The scan range available depends very much on the mechanical and optical design of your laser. Using a standard laser diode for cd-rom writing applications, without special coatings, and a low-efficiency 1800l/mm grating, scans of a few GHz are typical in a Littrow configuration. Increased scan range is possible by careful design of the pivot point of the grating, so that the frequency change due to the change in cavity length closely matches the frequency change due to the change in grating angle. See for example our paper Appl. Opt., 48 6692-6700 (2010).
We can instead ramp the diode injection current and cavity length (piezo) together. With standard uncoated diode, low-feedback grating and Littrow configuration, we can scan up to 40GHz without mode-hops. The figue below shows a 10GHz saturated absorption spectrum for the 780nm rubidium transition, with all hyperfine transitions for both naturally occurring isotope, together with locking error signal, using a standard MOGbox.
The frequency of a 671nm cateye laser (CEL002) was measured over several days as the room temperature and air pressure varied. The data below shows a 24 hour period in which the temperature changed by more than 2°C and the air pressure by several mbar. The laser frequency was within a band of ± 100 MHz, i.e. around 100 MHz/°C.