Coaxial cables are the most-common and vital conduits of RF energy, that's not news. If you flip through any RF-centric design publication or web site, you'll see that a large fraction of the advertisements and articles are about coaxial cables of many types. There are versions optimized for frequency range (a few GHz to tens of GHz), power-handling rating, physical ruggedness, temperature coefficient (causes phase shifts), matched performance, bendability, and other critical parameters.
But there's a medium for electromagnetic energy that is starting to compete with coax: optical fiber. A recent article in Microwaves & RF , Why More RF Engineers Are Choosing Fiber Connectors, explained the whys and hows of using fiber for both short distance as well as in-building longer runs, in small-cell applications and distributed-antenna systems (DAS). The much-lower attenuation of fiber (around 1 dB/km) compared to coax (roughly tens of dB/km) makes it especially attractive in the latter cases. Fiber's immunity to EMI/RFI is a big plus, too, especially in wide-area installations that have to deal with building motors and equipment, well-known and notorious interference sources.
Key to using fiber is the RF-to-optical (aka electro-optical, or E/O) converter at the source end and the complementary optical-to-RF converter at the receiver. For the transmitter, a distributed-feedback (DFB) semiconductor laser is used when wide dynamic range and low noise are critical, while for applications with lower-performance requirements, a Fabry-Perot (FP) laser is generally chosen. At the receiver, a PIN diode captures the photons and coverts them into electrical signals.
There's an element of irony here: although we associate fiber-optic links with high-speed digital signals, as in the Gbit/sec cables and links which are the foundation underpinning the physical layer of the Internet and its data flow, this use of optical fiber for RF is entirely in the analog domain –as so much of RF still is. Thus, the traditional analog issues of noise, linearity, distortion, clipping, limiting, and similar play their usual roles. Once again, analog circuitry and functions are necessary and unavoidable, and the optical drivers and receivers need to be optimized for their analog-performance and parameters rather than the digital ones (although both have the same root issues, as defined by the laws of physics, of course). Goodbye to eye-pattern woes, hello again to linearity headaches.
Of course, using fiber is not a trivial exercise. For engineers and installers whose experience and expertise are only with coax, there's a new world of optical-cable specifications (the fiber plus the protective jacketing), optical connectors, bend radius, E/O converters and components, and more. Plus, it's often makes sense to run a fiber-optic cable assembly which has unused fibers inside for additional capacity in the future (known as “dark fiber”); in contrast, running extra coax in parallel with the in-use coax is much more costly and occupies much more space.
Have you ever wished you had a performance- and cost-effective alternative to coaxial cable? Have you explored using analog fiber for RF? What concerns and issues would you have from a technical as well as personal standpoint?