International CES - Inside the Chassis of CES
Keeping it Real
To me so far -- and I've only seen a small fraction of the exhibits -- this year's Consumer Electronics Show has been one mainly of trends and imminent breakthroughs. Only a few actual, available products have piqued my engineering interest compared to the number of product prototypes and promising and truly exciting near-future technologies I've seen. One of the more intriguing and real, available-for-purchase products I've come across was the Panasonic SA-XR10 ($599, July). Compared to everything else that was happening in the truly massive Panasonic booth, this device, a multichannel receiver, was little heralded. It didn't even merit a picture in the Panasonic press kit CD-ROM (I took our picture). But its rated 5 x 100 watts per channel power output from an enclosure little bigger than a laptop computer was surprising. This was possible due to its use of extremely electrically efficient "1-bit" digital amplifiers (controlled by Texas Instruments chips), which enable high output power ratings with neither a bulky and heavy power supply nor with massive and expensive heat sinks. One-bit amplifier technology seems to be taking off in Japan, where an industry consortium has been formed to promote its use. We are at the beginning stages of a one-bit onslaught.
| Yamaha 1-bit power amplifier prototype |
Yamaha was showing a prototype of a one-bit power amplifier and Sharp was again exhibiting its one-bit technology that's now a couple of years old. Sharp's twist this time was that they were feeding, through a proprietary and copy-protected interface, one-bit data from an SACD player directly into their amps. The SACD data didn't drive the speakers directly, however, but had to be converted to the Sharp output-modulation scheme first. I was amused -- and annoyed -- by the presenter in Sharp's booth who restated the old canard that the 1-bit resampling rates employed (5.6 MHz, 128 times the CD sampling rate) ensure a more accurate reproduction of the signal (which, on the contrary, is limited by the original sampling rate). His talk of "limitless fidelity" was definitely overstating the case as was his showing of the now-standard -- and thoroughly misleading -- waveform graph showing how many more "data" points are handled by the high sampling rate. Someday I'll take such cant up in my column.