You may have seen recent reports that ultrawideband (UWB) technology is having a tough time, "Report: Ultrawideband dies by 2013." Some of its IC proponents have folded or are in dire straits, and one analyst flatly states that it will be dead in a few years. At the same time, higher-speed and longer-range versions of Wi-Fi, such as WiMAX, seem to be getting traction, "Wi-Fi backers drive to 60 GHz."

I'll be the first to say that I don't place much stock in what the pundits say; they are no smarter than most, and have no better crystal ball than any of us, IMO. But there is no doubt that UWB is not getting industry support or design-ins as hoped for by its proponents.

Why is this? Again, pundits (including me) always have excellent rear-view vision, and it's all too easy to glibly explain what happened and why; if only our foresight were as accurate.

But my view is that UWB suffers from two problems: it's not clear what problem it solves that can't be solved by other, perhaps less expensive or difficult approaches; and it is largely a new standard rather than an extension of an existing one.

One thing that helps a new standard take hold, grow, and really insinuate itself into the industry's fabric and end-user's world is when it is an extension of an existing, successful standard. This means that successive, improved generations can be backward compatible, a big plus in the real world of products and consumers. IEEE 802.xxx and USB are good examples of this approach. Or consider the much more mundane, basic-landline telephone. This is a standard that has served us well for over 100 years, and nearly every new feature or enhancement (Touch-Tone in place of pulse dialing, Caller ID, Call Forwarding, Call Waiting) did not obsolete previous installations, but built on them with continuing support and compatibility at the phone company's central office and local loop.

A substantially new standard requires significant investment in infrastructure, design know-how, and user experience and acceptance. And while an existing standard eventually runs out of extensibility (in plain terms, it runs out of gas) and must be replaced or superseded (think IP-based phones replacing the older standard), it takes a compelling argument to make it happen.

Follow-up on last week's column: Many of you responded to last week's column "In recognition of the older discrete component" with you own list of components with great longevity, in addition to the 2N2222 transistor I cited. Among them were the 2N3055 (along with its companion, the MJE2955), BU208, BC547, BC548, uA741 op-amp or many of its derivatives, 1N4148, IN4007, LM555,2N2905 PNP, 2N2907 PNP, LM317 positive regulator, LM337 negative regulator, 2N7000 FET, 1N914 small-signal switching diode, 1N4000 series rectifier, 1N5817 Schottky, 6N137 optoisolators, LF411 JFET op amp, TL072 dual JFET op amp, and the CLC436 high-speed op amp. ♦

Bill Schweber, Site Editor
bschweber@techinsights.com