Similar to coaxial cable transmission systems, optical network systems also require optical signal coupling, branching, and distribution, which necessitates optical splitters. An optical splitter, also known as a fiber optic splitter, is one of the most important passive devices in a fiber optic link. It is a fiber optic connector with multiple inputs and multiple outputs, commonly represented as M×N, indicating that a splitter has M inputs and N outputs. In fiber optic CATV systems, optical splitters are typically 1×2, 1×3, or 1×N splitters composed of these.
1. The splitting principle of an optical splitter
Optical splitters can be classified into two types based on their operating principle: fused biconical taper (FBT) and planar waveguide (BBT). FBT products are formed by fusion splicing two or more optical fibers sideways. Planar waveguides are micro-optical element products, using photolithography to form optical waveguides on a dielectric or semiconductor substrate to achieve branching and distribution functions. These two types share similar splitting principles; they achieve different branching amounts by changing the evanescent field coupling between fibers (coupling degree, coupling length) and by changing the fiber radius. Conversely, they can also combine multiple optical signals into a single signal, called a combiner. FBT fiber couplers have become the mainstream manufacturing technology in the market due to their simple manufacturing method, low cost, ease of integration with external optical fibers, and resistance to mechanical vibration and temperature changes.
The fused biconical taper method involves bringing two (or more) uncoated optical fibers together in a specific manner, heating them at high temperatures until they melt, and simultaneously stretching them to both sides. This results in a special waveguide structure in the heated area, forming a biconical shape. By controlling the angle of fiber twist and the length of stretching, different splitting ratios can be obtained. Finally, the tapered area is cured onto a quartz substrate with adhesive and inserted into a stainless steel tube, thus creating an optical splitter. This manufacturing process suffers from a problem because the coefficient of thermal expansion of the adhesive differs from that of the quartz substrate and the stainless steel tube. This difference in thermal expansion and contraction due to temperature variations easily leads to damage to the optical splitter, especially when placed outdoors. This is the primary reason for the susceptibility of optical splitters to damage. For splitters with more channels, multiple two-split units can be used.
2. Common Technical Specifications of Optical Splitters
(1) Insertion loss.
The insertion loss of an optical splitter refers to the dB loss of each output relative to the input. Its mathematical expression is: Ai = -10lgPouti/Pin, where Ai is the insertion loss of the i-th output port; Pouti is the optical power of the i-th output port; and Pin is the optical power value of the input port.
(2) Additional losses.
Additional loss is defined as the sum of optical power lost in dB relative to the input optical power across all output ports. It's worth noting that for fiber optic couplers, additional loss is an indicator of manufacturing quality, reflecting inherent losses during the fabrication process. Lower losses are better, serving as a key metric for quality control. Insertion loss, on the other hand, only indicates the output power at each port, taking into account not only inherent losses but also the splitting ratio. Therefore, differences in insertion loss between different fiber optic couplers do not necessarily reflect the quality of their fabrication.
(3) Spectrophotometry.
The splitting ratio is defined as the ratio of the output power of each output port of an optical splitter. In system applications, the splitting ratio is determined based on the actual optical power required by the optical nodes in the system (except for average distribution). The splitting ratio of an optical splitter is related to the wavelength of the transmitted light. For example, when transmitting 1.31-micron light, the splitting ratio of the two output ports of an optical splitter is 50:50; when transmitting 1.5-micron light, it becomes 70:30 (this is because optical splitters have a certain bandwidth, that is, the bandwidth of the transmitted optical signal when the splitting ratio remains basically constant). Therefore, the wavelength must be specified when ordering an optical splitter.
(4) Isolation degree.
Isolation refers to the ability of an optical splitter to isolate optical signals from other optical paths. Among the various metrics mentioned above, isolation is of paramount importance for optical splitters. In practical system applications, devices with an isolation level of 40dB or higher are often required; otherwise, the performance of the entire system will be affected.
In addition, the stability of optical splitters is also an important indicator. Stability refers to the ability of an optical splitter to maintain a relatively constant splitting ratio and other performance indicators despite changes in external temperature and the operating status of other components. In reality, the stability of an optical splitter depends entirely on the manufacturer's manufacturing process; products from different manufacturers vary significantly in quality. In practical applications, I have indeed encountered many low-quality optical splitters that not only degrade rapidly in performance but also have a very high failure rate. As a crucial component in fiber optic trunk lines, careful consideration is essential when purchasing them; price should not be the sole factor, as optical splitters with lower manufacturing processes are naturally cheaper.
In addition, uniformity, return loss, directivity, and PDL all occupy a very important position in the performance indicators of optical splitters.
Currently, optical splitters mainly use two types of technology: planar waveguide technology and fused biconical tapered technology.
1. Planar waveguide type optical splitter
PLC consists of an optical splitter chip and fiber arrays coupled at both ends. It adopts semiconductor technology, has good process stability and consistency, loss is independent of optical wavelength, good channel uniformity, compact structure and small size, and mature technology for large-scale industrialization.
2. Fused tapered fiber optic splitter
The fused fiber tapering technique involves bundling two or more optical fibers together and then melting and stretching them on a tapering machine. One end retains one fiber (the rest are cut off) as the input end, while the other end serves as a multi-output end.
3. Comparison of the performance of the two devices a) Operating wavelength
Planar waveguide optical splitters operate at wavelengths from 1260 to 1650 nm, covering the wavelengths required for various PONs at present. Tapered optical splitters can adjust the wavelength to 1310 nm, 1490 nm, 1550 nm, etc., but the manufacturing process is more complex and difficult to control; insertion loss changes with operating time and temperature. b) Splitting Uniformity: Planar waveguide devices have excellent splitting ratio uniformity due to the high consistency of semiconductor processes. Tapered splitters have poor splitting ratio uniformity, but their variable splitting ratio is their biggest advantage. c) Temperature Dependence (TDL):
Planar waveguide devices exhibit relatively small temperature variations; tapered splitters show significant temperature variations in insertion loss. d) Cost: Based on current production costs, 1×8 is the critical point; PLCs of 1×16 and above offer significantly better cost-effectiveness, while tapered splitters of 1×4 and below offer better cost-effectiveness. e) Reliability: Compared to tapered splitters, PLCs theoretically have only two fault points at their interface, while a 1×N tapered splitter has 2N-3 fault points.
