What is the Structure and Working Principle of WDM Devices?

There are many types of WDM (wavelength division multiplexing) technologies known at present, such as FBT (Fused Biconical Taper), FBG (Fiber Bragg Grating), TFF (Thin Film Filter), AWG (Arrayed Waveguide Grating), EDG (Etched Diffraction Grating), MZI (Mach-Zehnder Interferometers), and MRR (Micro Ring Resonator). Among them, TFF and AWG are the two most commonly used WDM technologies. This article will introduce the structure and working principle of TFF WDM devices.

The structure of a three-port WDM device includes a dual-fiber collimator, a single-fiber collimator, and a TFF filter.
 
The TFF (Thin-Film-Filter) is the most essential and expensive component in the entire WDM device. The TFF filter is attached to the end face of the collimating lens of the dual-fiber collimator, and its main function is to transmit and reflect signals.
 
The standard size of the film filter is shown in the figure, but there are also special sizes available. The film filter has a filtering surface (reflection surface) and a transmissive surface, with the main function of the filtering surface being to allow light of a certain color (corresponding to a certain wavelength) to pass through while reflecting other colors of light. The main function of the transmissive surface is to let light pass through the film filter. Generally, the reflection surface has a darker color, while the transmissive surface has a lighter color.
 
Multi-layer dielectric film filter is a type of multi-layer high-reflection film. The number of film layers can reach dozens or even hundreds, alternately composed of two dielectric materials with higher and lower refractive indices. The filter substrate Layers adjacent to air have a higher index of refraction. Combining dozens of layers of different dielectric films to form an interference filter with specific wavelength selection characteristics can achieve the effect of separating or combining different wavelengths.
 

The light emitted from the fiber end is divergent, so it cannot be transmitted far! What should be done?
 
Therefore, a collimator is needed to gather the originally divergent light into a parallel beam with a large spot, so as to achieve the collimation (parallel) effect and ensure a relatively long transmission distance. The collimator uses the converging principle of the lens (C-Lens or G-Lens).
 
What can make the diverging rays transmit in parallel and make the parallel rays converge? That would be the lens. The lens is a key component for transforming light beams, and the most commonly used ones are lenses with fixed refractive indices, which are also known as conventional lens(C-lens), and self-focusing lenses, also called gradient-index lenses (G-lens). Both C-lens and G-lens have focusing and imaging capabilities. The roles of the two lenses are different - the first lens makes divergent light rays parallel, and the second lens converges the parallel light rays.
 
One end of the C-lens is a spherical surface, and one end of the G-lens is a flat. It is this characteristic that allows optical components to be directly bonded to the flat end of the G-lens collimator, resulting in a more compact module, a feature that the C-lens does not possess.One reason why a G-lens is used at the input end of a WDM device is because its coupling face is flat, making it easier to affix a filter.
 
An important parameter in the collimation characteristics of a GRIN lens is the pitch. As shown in the diagram below, when a beam of parallel light enters a GRIN lens, its propagation trajectory follows a periodic function pattern. When the thickness of the GRIN lens is exactly one cycle, the exit light is also a group of parallel light rays. In fiber optic communication, a G-lens with 1/4 pitch is typically used.
 
By mounting a C-lens in front of a fiber optic head and encapsulating it with a glass or metal sleeve, a C-lens collimator can be created. The fiber collimator is formed by precise positioning of the pigtail and lens, using the convergence principle of the lens (C-Lens or G-Lens) to converge divergent light into a parallel beam with a larger spot size, achieving a collimation (parallel) effect. Generally, the cost of a G-lens collimator is higher than that of a C-lens collimator, so we mostly use C-lens collimators.
 
In TFF WDM devices, the dual-fiber collimator at the input end generally adopts a G-lens lens collimator, and the single-fiber collimator at the output end adopts a C-lens lens collimator.
 
Regardless of the package form, the basic optical path of a Filter-based WDM device is shown in the figure below. WDM signals include wavelengths λ1, λ2,...λn, which are input from the common end. The TFF filter allows one wavelength λn to be transmitted, while other wavelengths are reflected, so the wavelength λn is output from the transmission section, while other wavelengths are output from the reflection end. Wherein, one input optical signal is divided into two different optical signal outputs, which is demultiplex; two input optical signals are synthesized into one mixed optical signal output, which is multiplex.
 
In order to demultiplex all wavelengths, n three-port devices need to be connected in series to form a WDM module, as shown in the figure, where the TFF filters in each three-port device have different transmission wavelengths. The WDM module can be used as a demultiplexer or a multiplexer, depending on the direction of signal transmission.