THE REPEATER PRINCIPLE

...

You may now forget that ancient but very serious specification among electronic engineers for the ideal signal amplifier: A straight wire with gain. It has happened! But elsewhere.

In all these years we haven't even remotely approached that kind of simplicity--and surely not in all our present, complexly layered chips and our incredible ICs and the travelling electrons and holes that race through them. But, even so, we have come astonishingly close to the ideal of no distortion. Just look at all those zeroes after the decimal point in current audio amp specs. It is an enormously complex miniature world, our present electronics, if highly "practical," to say the least. But no straight wire.

What is now in existence, after the usual many years of dormancy for such profound developments, is a straight fiber with gain. Our friends in the glass business have done it. A tiny glass thread doped with a small collection of rare-earth ions, the most favorable being those of a substance that is new to me--erbium. No, not an herb! You do not grow erbium in your home erbarium. Think in another direction, of such now-familiar substances as germanium, gallium, indium, the very basics for modern electronics, and thus you will fit erbium into your consciousness.

It was some 20 years ago that the amplification principle was discovered-as usual, ahead of its time, a sort of curiosity. Dope a glass fiber with your choice of rare-earth ions (others also exhibit the effect), "bombard" them with a few tiny milliwatts of laser power, and lo! Amplification. The ions are pumped up a state or two, and as they fall down, they emit a more powerful version of whatever light signal is passing by. Pardon my non-engineering approach.

I picked up my first account of this extraordinary phenomenon from that useful reporting vehicle, Scientific American, back in March 1991. I left it out of my October 1991 column because I had instantly realized that this astonishing development needed its own space, plus a bit of background.

The straight-fiber amplifier is, you see, essentially a repeater, inserted in a long optical-fiber channel to boost up the light signal as it begins to fade. The digital aspect of fiber optics means that extraneous material, as always, can be ignored for the essential bits, and even missing bits can be restored by methods becoming more familiar to us every day. The repeater is simply the first use for a straight-fiber, laser pumped light amplifier, and for a good while it will be overwhelmingly the most important. You will remember, as of October's Audio, that in the mid-'80s a single-throw distance of some 117 kilometers had been achieved with fiber transmission, a notable advance over the earlier "long" distances, across town, achieved by the telephone company. But 117 kilometers (about 73 miles) is not a real long distance. So repeaters were quickly developed to extend fiber cables for enormous distances-but, as I understand it, entirely via an unhappy substitute for the real thing, a sort of A-to-D and D-to-A conversion, from light into electronics and then, amplified, back into light. Apparently (I have not pulled up any glass fibers to see for myself) the entire present world fiber network depends on this workable but highly unsatisfactory procedure.

The erbium-laser amplifier is still mostly prototype, and some absolutely vital standards are not yet decided on.

Alas, at this point the optimum laser is incompatible with the optimum tailored digital fiber, the bit rates, and much, much more, as the ads say. Do we rip up thousands of miles of glass communications with electronic repeaters and substitute all-new erbium-doped fibers throughout? That is one horrendous possibility. Do we compromise--the old, old game of compatibility at a sacrifice? Or do we sweat it out until some ingenious guy or corporate team discovers the perfect way to adapt an optical amplifier to existing fiber? Let us hope so. Evidently all sorts of important outfits, the world over, are plugging hard; it is a vital matter for future communications. We probably will not see a useful resolution for a settled worldwide fiber network, all digital, all optical, until the middle of this decade.

Note that the repeater principle has an ancestry that goes back to the beginning of the human race, whereas plain amplification, a small effort, a small signal, into a much larger facsimile, is a more obscure and much later idea. Think of the relay race; think of the ancient human messenger, running full speed until he totters half-dead as a new and fresh messenger takes over; think of the smoke-signal repeaters, by which hideous emergencies could be made known across hundred of miles in remarkably short order. Then there was the stagecoach or the hired private coach, dashing ahead at full horsepower, and the highly perfected change of horses where, in early 19th-century England, a complete new "motive power" could be installed on a large coach in literally seconds, scarcely a pause at all. Human and animal repeaters have always been with us, as essential parts of communication.

Think also of the telegraph, the first true electrical means of communication and far ahead of all the rest, back in the 1840s. Morse had the first digital system--can anyone question it? Two symbols (and only one transmission wire) coded into the entire language.

But the repeater was a very tough problem. The first Morse line, between Washington, D.C. and Baltimore, got by without a repeater, if I am right--but barely. It still created a sensation, just as had the original discovery that electricity moved instantaneously. But the telegraph that followed the railroads across our continent and everywhere else was in dire dependence on some sort of repeater to keep the signal going. All too often it was human. A fading message was written down, as much as could be deciphered, and then tapped out anew on a fresh circuit. After dozens of such transcriptions, most telegrams were reduced to jargon! Like the children's game where a few words are whispered ear to ear around a large circle, the message ending up in chaos.

The true repeater solved much of the jargon problem--it was an electrical amplifier that actually was mechanical.

A feeble make-and-break actuating a much stronger, fresher current. An electronic repeater, which had to involve amplification, was not devised for three-quarters of a century, by then applying also to the telephone.

The digital nature of the telegraph gave it a tremendous boost-retrieve the "bits," and you had the signal but not the noise. You can see, thus, that the phone network began with a very serious handicap, which is probably why, when I was a child, a long-distance call was for some extreme emergency and a clumsy thing at best. I still tend to talk loud when I phone across the country. We all did.

I am no authority on the inner details of the new light-amp repeater-amplifier, and I refer you to the Scientific American source. In non-tech language, I can give you a rough idea as to how it works. The pumping laser, whatever type is eventually perfected, is hearteningly similar to our ubiquitous CD lasers and their relatives. Not big, very modest in power requirements, not expensive-especially for professional communications networks. As I understand it, the signal fiber is joined at its doped area by a second fiber, which carries a diode-laser pulse that makes the erbium amplify. The doping may be concentrated in a small loop or distributed over a long distance. The erbium (or other similar material) emits light in the familiar way, imposing its new power on the light signal that is passing through. Lots of ifs and buts in that, I am aware. But I assure you it exists. It will be perfected. And multiplied. Moreover, its use will soon expand, without any doubt at all. Into audio? Extremely likely.

There is a lot more than just amplification involved in this new way to repeat an optical signal. As we know, light has an enormous bandwidth. For our purposes it might as well be infinite, a million or so separate digital "channels" imposed on a single beam of light! (See October.) Our own electronics have explored the furthest reaches of our kinds of bandwidth, but we can never compete with light. What, then, do we do when a wideband, multi-channel light signal fades and heads into an electronic repeater system? For lack of bandwidth, we must sort out the channels and amplify them separately. We are limited in all sorts of ways by this-can you imagine a repeater that needs a thousand amplifiers for an easy and casual thousand channels carried in the light signal? Also much associated clean-up equipment and assorted filters--you know what filters get you into--both to separate and to recombine the numerous electronic channels. And then, of course, the repeat conversion back to a single light-beam signal. These are the light repeaters now in use, and they are roughly equivalent to the mechanical repeaters of the early telegraph. What a bottleneck! By the way, just as the erbium ions can be doped in concentrated clumps at one point in a transmitting fiber, or spaced out thinly for a sort of gradual amplification over considerable distances, so too can the pumping laser adapt to conditions. Its light-conducting fiber can couple with the signal fiber perhaps only inches away. Or if your repeater happens to be under the English Channel or at the bottom of an ocean 30,000 feet deep, you can set up your laser on the nearest convenient land and run a second parallel fiber alongside the first, two human hairs together, to carry the laser beam to the point of junction and amplification. How flexible can you get? And then there is the solution, a "wave-like optical pulse" that will go 10,000 kilometers without spreading out... . The Japanese, according to the Scientific American article, plan a brand new, 2.4-gigabit all-optical network under the Pacific, to open up in 1996.

Details not yet settled-they aren't settled anywhere. As to audio itself, the first thing to note is how much of this new technique meshes into existing knowledge and practice of many sorts and definitely into our own ever more profound concern with semiconductors and the laser, in particular, so familiar in audio CD players. How can we possibly NOT get involved? But in new ways, not just repeaters.

I suspect erbium amping will show up first in our near neighborhood, the video cable, already carrying plenty of audio, feeding most of our equipment with its signal. Have you seen those fat things that now bring cable into a million homes? They are an expensive vested interest, but new territories await the cable that is hair-like and can carry absolutely anything, with repeaters built in. My town does not have cable-not enough houses. If our local outfit has heard of erbium, they may merely be stalling until it arrives. If not, sooner or later they will discover it. A cable revolution again, maybe the end of networks, maybe even radio, going along on cable for the ride? Well, big changes anyhow.

In specific audio equipment for the home and car, I can only give a guess that erbium-light amplification would be most useful in the low-level areas, the delicate and critical preamp and its numerous relatives where simplicity and purity are of the essence. Right? The straight wire turned fiber.

(by: EDWARD TATNALL CANBY; adapted from Audio magazine, Jan. 1992)

= = = =