Cavity Design

There are two main types of fiber laser cavities: Fabry-Perot and Ring cavities. Of these two basic types, there are many variations in their specific implementations. We will describe those variations under the next two sections.

Table of Contents:

  1. Fabry-Perot Cavities
  2. Ring Cavities


We discussed Fabry-Perot cavities (interferometers) in the “Classical Optics” portion of this website, but here we will discuss how this would work with fiber lasers. The most basic implementation, and indeed the implementation used in many early fiber laser experiments, is formed by placing the fiber gain medium between two mirrors. The fiber ends would in fact normally be butt-coupled to the mirrors to reduce losses. The mirrors would be designed to be highly transmissive to the pump wavelength but highly reflective to the signal wavelength; thus the cavity is formed. However, as one may imagine if you have already read through the description of the Fabry-Perot cavity in the “Interferometry” section, any slight misalignment of either the fiber end or the mirrors would lead to non-negligible losses of the signal. This is because the coupling and resonance condition is highly dependent on the incident angle. The problem was partially solved by depositing the partially reflective, partially transmissive dielectric coatings directly on the ends of the fibers. However, there are problems with this design as well because the coatings are quite sensitive to any imperfections on the fiber tip and can even become damaged with the high-power, focused beam of the pump.

To solve this problem, several solutions have been explored and used in practice. One method to prevent the pump light from going through the dielectric mirrors is to use a WDM, or wavelength-division multiplexer, to couple in and out most of the light before it reaches the dielectric surfaces. Another method is to etch fiber Bragg gratings directly onto the core of each fiber end using holographic techniques. A setup with this method is shown in the figure below.

OSA | Effective length of short Fabry-Perot cavity formed by uniform fiber  Bragg gratings
Taken from: Yuri O. Barmenkov, Dobryna Zalvidea, Salvador Torres-Peiró, Jose L. Cruz, and Miguel V. Andrés, “Effective length of short Fabry-Perot cavity formed by uniform fiber Bragg gratings,” Opt. Express 14, 6394-6399 (2006)

The principle of operation of this method of providing feedback is that the Bragg gratings can be selectively designed to be transmissive to the pump but reflective for the signal. As we learned in the “Light Propagation in Crystals” section, the Bragg condition is dependent on wavelength, so Bragg gratings can be used to totally reflect wavelengths that match the Bragg condition. All other wavelengths would be hardly affected by the Bragg grating, except potentially for some spectral sidelobes depending on the spectrum of the pump. The Bragg condition for a fiber Bragg grating is given as:

\begin{equation} \lambda = 2n_{eff}\Lambda \end{equation}

where “$\lambda$” is the signal wavelength, “$\Lambda$” is the periodic perturbation of the refractive index to create the Bragg grating, and “$n_{eff}$ is the effective refractive index seen by the light passing through the fiber. Notice that there is no angular dependence in this definition of the Bragg condition, which arises because the Bragg grating is etched directly into the fiber. Essentially, then, for light of wavelengths that satisfy this condition, they will be reflected back the same direction as the incident wave. The reason a grating is needed is because the amount of light reflected back depends on how much the refractive index is modulated, so it may take a few “tries” before all of the light is reflected back.

Another option is linear loop mirrors, which basically just act as a Sagnac interferometer that can transmit the pump wavelength but reflect the signal wavelength. The way this works is by using a fiber coupler where the two output ports are spliced together. Pump light sent into one end of the fiber loop mirror will then be split into two counter propagating directions; depending on the interference when they meet back at the coupler, different amount of light will be transmitted into the fiber and sent back towards the pump.


One of the great advantages of ring cavities is that they can be made without using mirrors, meaning that you can have an all-fiber laser. In one of the simplest designs, a WDM coupler is used to couple the pump into the laser ring, and then the output light emerges from the other port of the coupler. An isolator is used to preserve unidirectional travel in the loop.

Another cavity design used more often for mode locked lasers is a figure-8 laser cavity, because it consists of two loops in a figure-8 shape. This type of cavity will be discussed in further detail in one of the later sections, but it involves the control of nonlinear effects.

%d bloggers like this: