How They'll Grow TV Sets Like Tomatoes
August 1961 Popular Science

August 1961 Popular Science [Table of Contents] Wax nostalgic about and learn from the history of early electronics. See articles from Popular Science, published 1872-2021. All copyrights hereby acknowledged.

Big plans were being made for solid state electronics by the time this "How They'll Grow TV Sets Like Tomatoes" article appeared in a 1961 issue of Popular Science magazine. The weird title alludes to "growing" integrated circuits (ICs) from crystals. Fairchild Electronics' Robert Noyce had demonstrated the world's first monolithic IC two years earlier, and rightly so, futurists were creating fantastic schemes for not just limited functionality IC like amplifiers and logic gates, but entire systems comprised of mixed signals (digital and analog) ranging in frequency from DC to light. Dr. Noyce died in 1990, so he had plenty of time to watch the explosive grown of the technology he co−invented. Unfortunately, he missed the smartphone and cellular telephony system build−out, the Internet, and millimeter wave / optical wavelength ICs, micro electro-mechanical systems(MEMS), integrated fluidic circuits, and much more in the ensuing two and a half decades. Interestingly, gallium arsenide (GaAs) is featured at a time when germanium and silicon were the majorly dominant semiconductors.

How They'll Grow TV Sets Like Tomatoes

New technique may create radio and TV receivers all in one piece - each a single block of crystal containing no separate parts

By Martin Mann

Of all the fantastic inventions of the guys who play around with transistors, the latest is the wildest and most wonderful. They are going to grow radios, TV sets, and other electronic gadgets as they grow tomatoes or carrots.

Their crop will be tiny, shiny specks of crystal. Each crystal will be the works of a radio, say. Not one part for a radio circuit, but the whole thing. There will be no parts - just a single solid crystal. You'll still need loudspeaker and knobs on the front, antenna and maybe batteries at the back - but in between only a crystal "radio block."

That solid block will do everything that now requires 50 separate parts: half a dozen transistors, dozens of resistors, condensers, and coils, and a maze of metal-ribbon connections. A radio signal will come in the back of the block, and audio power for the loudspeaker will come out the front. It will be so small it will look like a rivet on the loud-speaker frame.

Does this sound as if some press agent has been belting his expense-account bourbon too hard? The fact is, the first crude versions of these miraculous buttons have already been made.

Last winter, Westinghouse engineers demonstrated a small radio. They couldn't quite make it in one block - not yet. They used eight. The Air Force has ordered a high-frequency communications receiver built the same way. And Texas Instruments, the transistor tycoons from Dallas, are already selling their Type 502 multi-vibrator. This gismo for a computer, so small you can barely read the trademark, does what formerly required 16 separate parts.

The big bonus: reliability. With just one part and few connections, an electronic block should work every time

It will be a couple of years before gadgets you can use show up in the appliance stores. But when they do, you'll know a real revolution has come. This revolution means:

• Handy size. TV sets might be no bigger than picture screens, and these could be flat as a book. The personal radio would shrink under wristwatch size and become part of the watch, controlling the timekeeper as well as furnishing music; it might need no battery, making its own electricity from the warmth of your wrist. Really smart electronic computers might become small and cheap enough to drive your car or control household appliances.

• Economy. The all-in-one blocks just naturally come in big quantities. You grow one giant crystal, then dice it; each bit could be an entire radio chassis or computer circuit or whatever. The process has to be automatic (no human can manage it), so it should be cheap - eventually.

• Long, trouble-free use. About the only thing likely to make a solid-block radio or TV go dead is a small boy with a hammer. A solid unit doesn't burn out; there are no interconnecting wires to come loose; and - most important - reducing the number of parts automatically cuts the chance that anything will go wrong.

• No repairs. You won't be able to fix one of these things because there will be nothing to fix. You'll throw it away, like a torn Dixie Cup. (Servicemen may be automated into the ranks of the technologically unemployed. Who needs an expert to replace faulty parts when there's only one part to replace?)

If this sounds good to you, it sounds even better to military men. What excites them most is reliability. Complexity is the reason many missiles go haywire. With so many parts, one or another is almost bound to fail. Reducing the number of parts is a much more effective route to sure-fire operation than trying to improve the parts - some individual components are already specified for 99.993-percent reliability (no more than seven hours of failure in 100,000 hours of operation).

The tremendous savings in size and weight that the solid blocks will bring is just as valuable. We'll be able to send up bigger loads with less powerful rockets.

Beyond that, an awesome prospect comes into view. These magic blocks of crystal come close to matching human brain cells. For the first time, it really seems possible to build an electronic computer that will be as smart as a man: a black box not much bigger than a man's head and able to do the same things as his head. Such a development alternately delights and scares the pants off the philosophers.

People who know something about electronics often miss the revolutionary character of these solid crystal blocks. They know that resistors retard electricity, that transistors boost electricity, that combinations of resistors and condensers cause electricity to surge back and forth. So they think that the solid blocks are just a new way to make extremely tiny resistors, condensers, transistors, etc., and cram them into one small package. That's not the idea.

Coming closer. Then they think that somebody has figured out how to treat different sections of a solid crystal so one section is a resistor, another section a condenser, another a transistor, and so on. You would still have the familiar old components but mapped out on one single crystal instead of sticking up separately on a circuit board. The crystal becomes a complete circuit instead of one part for a circuit. That's closer. Texas Instruments uses this technique for its solid circuit multivibrator.

Yet even that is not the full flowering of a radical and beautiful idea. Forget about resistors, condensers, transistors, and even about circuits. There aren't any. All you have is a shiny gray block of unbelievably pure silicon or germanium into which have been squirted a few atoms of tin, zinc, or phosphorus. This block is designed from the beginning to perform a "function."

If it's a radio block, it takes in (from the antenna) a very weak current that varies in a complicated way and gives out (to the speaker) a powerful current that varies in a related, but simpler, way. You can't point to the block and say that the RF amplifier is here, the detector there, and the audio amplifier some place else. They are spread through the block. If you slice the block in two, you have two radios, each less powerful than the original. You can keep right on slicing, cutting off additional radios of less and less power. (There is a limit, of course; you end up with slices containing too few of the atoms needed to perform the radio function.)

For other functions, you don't rearrange the radio block. You grow a different block to suit the new function. A TV block, for instance, would have to handle more complicated currents, at higher frequencies, and deliver two outputs (sound and picture).

Arranging atoms. The making of solid function blocks is a triumph of molecular engineering, closer to chemistry than electronics. The designer cannot sketch a circuit made up of old-fashioned components. He computes the flow of electricity through the block, figuring what changes - in voltage, current, frequency, and so on - must be accomplished to perform the job. Then he must arrange atoms and molecules within the block in such a way that they will accomplish those changes.

Manufacturing function blocks is a sort of synthetic agriculture. The trade jargon borrows farmers' words. The process starts with a seed, a small but perfectly formed crystal of the purest silicon ever known. The seed is dipped into a pot of molten silicon, then - under the most precise control - pulled out. The seed grows into a big, perfect crystal as atoms from the melt freeze onto it in exactly the right places.

New crystal-pulling techniques grow "dendrites" - very thin and narrow, absolutely perfect, and almost as long as you want. These ribbons are just sliced up for use.

A pure crystal is useless; it won't even conduct electricity let alone perform a radio job.

It must be carefully doped: purposely dirtied with just the right amount of impurity (a few atoms per million crystal atoms) in exactly the right places within the crystal. A mistake of millionths of an inch wrecks everything.

This can be done while the crystal is being grown. Impurity pills are dropped into the melt, like fertilizer in a furrow, to introduce the essential ingredients in the right places. Ultimately, complete radio or TV blocks might be grown this way, coming out of the pot all ready to be sliced off, slipped into cabinets, and plugged in. Today only a few simple function blocks grow from the melt in their final form. Most require treatment after the crystal is grown. Impurity layers are put into the crystal by elaborate alloying, diffusion, and etching techniques.

These processes are delicate. Today it is difficult to produce a whole batch to meet specifications. Only simple types have been made - single-purpose blocks such as amplifiers. No one has yet been able to combine several different functions into one block, as you would have to do for a solid-block radio or TV. (Some doubting Thomases think that won't be managed for a long, long time, if ever.)

And the units are still expensive. Officers at Wright Air Development Division like to show off the "Air Force jewels" -a plush-lined leather case containing the first function blocks, which cost the Government far more than diamonds.

Yet the spectacular success that has been achieved suggests an even more remarkable future. Why stop with integrating the inner electrical works into one solid block? Why not go on from there and make the whole thing, front to back, a single unit?

Self-powered? Batteries or power cords would not be needed. Semiconductor crystals, doped the right way, will generate electricity from other energy (heat, light, or a hard squeeze). So function blocks could be their own power sources, making automatically all the juice they'd need to operate.

Radios need loudspeakers to convert electricity into sound. But some crystals (piezoelectric types, like those in a phonograph pickup) can do that, too. Such a crystal might be amalgamated into the radio function block.

TV sets need picture tubes to convert electricity into light. Solid crystals to do that job are apparently just around the corner; eventually one might become the front surface of a television function block.

Such gadgets would be the ultimate in electronics: complete, self-contained "things," like apples or tomatoes. Lacking wires, batteries, identifiable parts, their electrical nature would disappear. Electrons jumping around inside their molecules would still make them work-basically the same thing that goes on inside apples and tomatoes.

Posted May 14, 2024