The Death of Copper
Network World recently ran a series on the 20 greatest “Coke vs. Pepsi“-like battles in the networking industry. Surprisingly, they did not include anything like “POTS vs VoIP” or “Copper vs Fiber.” Perhaps it is because they, like me, don’t think the debate is relevant: VoIP and fiber have already won, but POTS and copper are never going away just as vacuum tubes never will (yes, there is still a need for vacuum tubes and always will be).
If we start by talking about copper at the physical layer of the network, there is no doubt that copper (Cu, 20 on the periodic table),a key component of the telecom industry’s copper wires, is a natural resource that is being depleted. Prices for copper increased 400 percent over the last decade. The cost of running a Cat 5/6 cable drop during that period went up proportionately, which was one of the main drivers behind more widespread WiFi deployment on corporate campuses. The convenience of Wi-Fi helped, but so did the fact that each wired jack in the wall cost four times as much as it used to.
Copper is still a brilliant final-mile or on-premise solution, regardless whether you are using POTS or VoIP at layer 2/3/4. It can take tight corners, unlike fiber, and it punches down on a demarc like a breeze. Furthermore, in a pinch I can twist-tie copper together, but my eyes still can’t quite laser-fuse a fiber splice (and believe me, there here been times when I have tried to channel Superman’s laser vision, to no avail). That’s not to say that cheaper plastic multimode fiber is not making its way further and further into the local network, as it has a few tricks up its sleeve that copper doesn’t. I will never forget my first engineering tip from US WEST technicians in 1991 about how to attenuate a fiber in a pinch: wrap it five or six times around your finger. Since light has a hard time bending, that trick quite effectively ratchets down the signal strength with each circumnavigation.
All this said, copper will be the mainstay for the final mile for a long time for one simple reason – it conducts electricity. This isn’t just because power over Ethernet is going to be the standard of the future (a good argument in itself). No, it is because people abandon their telecom infrastructure every time they move, and for some reason just don’t think labeling their cable runs for the next guy is that important! Ooops, oh …. can I withdraw that exclamation point? I feel it’s a little like the pot calling the kettle black, despite my best intentions. I just sold my house in Minneapolis and wrote on the disclosure, “All Coax, Cat 5, and audio-visual cabling transferred as is.” I honestly tried my best to keep it labeled and orderly and functional over the years, but with all the work associated with moving out – and all the wires running across/through/inside (insert description of possible code violation here) – I really wanted to wash my hands of it. Which has been my experience on every building I have ever taken over to install a network for a new tenant too.
So, what is the other reason copper’s conductivity will keep it king of the final mile? Signal tracing. Have you ever tried to co-opt the former copper infrastructure of a previous tenant (or several previous tenants) in a building? You would think all jacks are labeled properly and all jacks must go back to one central hub or punchdown panel, right? Wrong. What you’ll come to find is that many new tenants, giving up on trying to figure out where the last tenant’s wires were running, will build their own overlay and completely abandon some of the perfectly good copper infrastructure. Some jacks go to one punchdown panel or hub, and others go stealth (presumably connected to an unknown location either in a neighboring suite that is no longer connected or another floor where some previous multi-floor tenant had its main office.)
As such, if you ever want to save some money and make use of an existing copper infrastructure which would now cost 400 percent more to overlay, you need to be able to trace the circuits with a tone generator, which requires our good old friend, the copper conductor. By injecting a pulsating 48 volts down a wire, you can trace it all across a building without ever cutting into a splice, just like you could use a metal detector to find a buried piece of copper in the ground. Once you find the ends of the runs and can reroute them physically, you can re appropriate our precious copper friends. One note of warning – if you are going to re-use copper infrastructure, there is a sneaky little difference in how Cat 3 or Cat 5 voice wiring can be terminated.
Twenty years ago, I learned how to punch down or crimp CAT3/5 voice wiring: RJ-61, with the pairs ordered as blue, orange, green, and brown. I never knew until I ran into a dead end while volunteering to move a small business phone system recently that while RJ-61 is now analogous with T568A, a newer standard known as T568B is preferred. The only difference between A and B is that the orange pair and green pair are swapped. It’s a minor difference, but the rationale was that as IP telephony started to come into its own, it made no sense having one jack in an office wired one way (for an analog/digital telephone) and another jack wired the other way (for computers or IP telephony). The moral of the story: thank you, copper; I never would have figured it out without my electrical signal tracer and wikipedia.
So is copper dead? At the physical layer, no, never. But what about POTS vs VoIP? Is copper dead there? Indeed it is, and I’ll argue why in my next post.
Email
Bookmark
Digg
Tweet This
SubscribeTags: copper, copper vs. fiber, fiber, POTS, POTS vs. VoIP, T568B, VoIP
