Narrow-linewidth and frequency stable lasers are nowadays a key components in order to reach high performances in many domains and applications where laser interferometry process is involved, like for LiDAR, SENSING, METROLOGY, QUANTUM TECHNOLOGIES, …
The best way to characterize the laser emission frequency is by being able to detect the laser frequency fluctuations, which is done with an OPTICAL FREQUENCY DISCRIMINATION scheme. At SILENTSYS, we propose plug and play OPTICAL FREQUENCY DISCRIMINATOR (OFD) in order to reach high performances laser frequency characterization.
We will see here how it is possible to get a lot of information using an OFD. Simple, fast, efficiently !
OPTICAL FREQUENCY DISCRIMINATION : DEFINITION
OPTICAL FREQUENCY DISCRIMINATION, or LASER FREQUENCY DISCRIMINATION is a technic that gives the frequency evolution in time domain of a single frequency under-test laser as we have presented in another article. So, these are methods that convert laser frequency emission into a readable signal, because, in contrast to electronic oscillators, a laser emits an electric field at very high frequency typically, around 200 THz for telecom lasers (1,5µm wavelength).
At SILENTSYS, we are also doing OPTICAL FREQUENCY DISCRIMINATION, that is directly integrated in few of our products like the OFD (Optical Frequency Discriminator) or the OFC (Optical Frequency Correlator). We provide OPTICAL FREQUENCY DISCRIMINATION that are small, fast, affordable and available from 400nm to 2200nm with a typical working range of 100nm around the central wavelength.
These are based on modified Michelson or Mach-Zehnder fiber interferometers, with a high vibrations / acoustics isolation and high-level temperature stabilization (at the micro-Kelvin level) and ultralow noise photo-detection in order to reach very good laser frequency discrimination precision.
These products make good candidates for laser frequency characterization, enabling Frequency Noise PSD measurement, Frequency Stability analysis, and Frequency Noise reduction by more than 60dB to reach Hz-level laser linewidth from a MHz one.
HOW DOES IT WORK ?
As the SILENTSYS OFD/OFC are based on a modified Michelson or Mach-Zehnder interferometers, so a 2-waves interferometric process, the result is a sine fringes pattern by the change of the input laser frequency (as presented in another article). At the middle of one fringe (the green zone on the Fig.3), the conversion between laser frequency fluctuations and output voltage is linear, so it is then possible to retrieve the laser frequency fluctuations by recording the voltage time trace of the OFD and by knowing 2 important parameters which are: the amplitude of the fringes (Vmax) and the OFD free spectral range (so the frequency distance between 2 fringes).
In the next figure, as the laser is going through 4 fringes by changing its frequency of 4 MHz, it means that the FSR is 1 MHz in this specific case. The OFD FSR is fixed by construction and can be adjusting at the order between 2MHz and 2GHz (other values possible on demand).
LASER EMISSION DIAGNOSTIC USING AN OFD
As the SILENTSYS OFD/OFC are able to give in real time a signal that is linked to the laser emission (frequency and power), we can actually learn a lot about the laser emissions from the time-trace characteristics / behavior, as we will show here. For example, just with a quick lock at the signal viewed on the oscilloscope (output voltage of the OFD) we can evaluate more than 10 specific cases, as the instantaneous linewidth of the laser, the slow frequency drift, the optical power / polarization drift, the monomodality, mode-hopes, frequency locking quality, …
Below we present different cases to show how useful an OFD can be for your experiment by better understanding the laser characteristics.
Case #1 – Laser instantaneous linewidth
In this case, the thickness of the curve is directly linked to the white frequency noise of the laser, around 0V so at the quadrature point. If the trace is thick, it means a high white frequency noise floor, and if the trace is thin, it means the opposite. However, it is important to not have low pass filter to misestimate this contribution. Of course, the thickness of the curve will also depend on the OFD-FSR.
Case #2 – Laser frequency modulation
In this case, the fluctuation of the curve will give an important information of the frequency modulation of the laser. If you can observe sine function (so scanning the fringes), it seems that the laser is highly modulated (at either fast or slow frequency). Actually, when you are in the CASE #1, with a thick curve, it is always good to zoom in the curve to see if there is not a sine function, meaning a fast laser frequency modulation. However, in the case of frequency modulation, the signal reach the maximum / minimum voltage, which is an important data to understand.
Case #3 – Laser frequency drift (ex. random walk)
In this case, the fluctuation of the curve will give an important information of the low frequency contribution of the laser. If you can observe a “random” scanning the fringes, it seems that the laser is highly affected by noise at slow frequencies, by for example thermal effects.
Case #4 – Laser frequency mode-hopping
In this case, the fluctuation of the curve will give an important information of the laser frequency jumps, also call mode-hops. This is easily visible and is not a good sign. It means that the laser is probably affected by back-reflexions and creates an external cavity very long and instable, with very low FSR (MHz domain). With this configuration, active laser frequency locking it not recommended and probably not robust. External cavity with back reflexions also causes “artificial” white frequency noise reduction that can underestimate the real laser frequency noise.
Case #5.1 – Laser polarization state
In this case, the fluctuation of the curve will give an important information of the laser polarization state. Usually, the OFD need linear polarization at the input and are not fast axis blocked. So, the polarization of the laser needs to be aligned to the slow axis of the PM fiber. However, if the polarization is not well aligned to the slow axis, this will cause interferences on the fringes, that we can observe easily by changing the OFD temperature. So, to properly use the OFD, the polarization should be linear with the highest extinction ratio (we recommend minimum 18dB) and on the slow axis. The use of a polarizer can help.
Case #5.2 – Laser power fluctuations
In this case, the fluctuation of the curve will give an important information of the laser power fluctuation. This can be very close to what is observed in the case 5,1. The best way is to measure at the same time the laser output power to be sure which phenoma it is. However, a laser power fluctuation can be problematic because it will affect the Vmax and –Vmax so it will affect the discrimination factor…
