The modern world — the electronic age as we know it — has a birthday coming up: December 16.
On that day in 1947, a couple of guys at Bell Laboratories, John Bardeen and Walter Brattain, did a really cool and profoundly important experiment. They made this odd-looking contraption, and it WORKED:
It’s estimated that over 10,000,000,000,000,000,000,000 of these (ten sextillion!) have been produced, making this humble snarl of wires the most manufactured device in history, by a very, very wide margin. In fact, there are now over 1,000 of these jobbies for every grain of sand on Earth.
Is it found-object folk art? Do they do root canals with it? Maybe it’s a sewing machine for scarecrows?
No, that strange little item is the world’s first working solid-state TRANSISTOR. Bardeen, Brattain, and their boss, William Shockley, would go on to win the 1956 Nobel Prize in Physics because of it.
There are a couple of truisms out there about the transistor:
“Transistors replaced vacuum tubes.” “A computer used to be the size of a room!”
Those are definitely correct, but to bring it right up to the present, let’s look at a really state-of-the-art chip from a brand new iPhone: the “A13 bionic” chip. One edge of this little square is about a half an inch, yet it’s loaded with 8.5 billion transistors.
If you had to construct this A13 chip from the very smallest vacuum tubes you could possibly find, each about the size of a beefy corn kernel, your resulting smart phone would need to be at least as large as the White House.
So a transistor can do exactly what a vacuum tube does, but in a much, MUCH smaller space, which means it uses far less energy as well.
But … what does a vacuum tube even DO, anyway?
Well, the vacuum tube was — no, is — a really useful invention, thought up by Sir John Ambrose Fleming in the early 20th century. I come to praise the vacuum tube, not to bury it. In fact, you almost certainly still have a working vacuum tube in your house, and I’ll tell you where at the end of the diary.
The most fundamental thing a vacuum tube does is to allow electric current to flow in only one direction. Devices that do that are called diodes. A diode of today looks like a nondescript little grain of rice, because it’s made of bits of semiconductor materials:
But a vacuum tube does the job in a much more physical, Dr. Frankenstein kind of way.
Inside the vacuum tube are two metal plates, both part of the same circuit but separated by empty space. Normally no current would flow in a situation like that. But here we force the issue by heating up one of those plates, hot enough that negatively-charged electrons literally “boil” off of it. See, that’s already something an evil genius would do in their basement.
As you know from kids’ toys or your wall clock, we can hook up a battery the “right” way or the “wrong” way. If we do it the “right” way with a vacuum tube, the heated plate will be the negative end, or the cathode. Over at the other end we have the positive anode, so electrons are attracted to it, and that’s where they’ll flow. This gives us our current. But if we hook up the battery the “wrong” way, we’ll be heating up the positive side, so even though electrons will still boil off of it, they’ll be attracted right back to it, and we won’t get any current.
But now for the REALLY great use for a vacuum tube: Let’s add a little mesh between the positive and negative plates and call that the “grid”. This is a little metal plate with holes, so electrons can easily get past it.
We’ll keep the battery hooked up the “right” way between the cathode and the anode, so we start out with a nice current. But now we can use a separate power source — and it doesn’t take much — to put an electrical charge onto that grid. If we put a positive charge onto it, as you see in the right half of the picture, that attracts electrons, and they have no trouble zooming right through the grid. But if we put a negative charge onto the grid (at left), we make it repel electrons. Then we’ll have a smaller current, or even no current at all. Charging up the grid is a great way to regulate the current, because it’s much faster than using a physical switch or heating and cooling the cathode.
See what we did here? We used a small current to control a big current.
That’s a bit of a superpower we’ve taken on. Because now one thing we can do is AMPLIFY. If we have a radio receiver, it’s going to pick up a pretty weak signal out of the air. If you turn your radio down to volume 0 and stick your head right into the speaker, you’ll hear a tinny squeak at best. But that little bitty signal is enough to charge the grid, and if we have a nice big battery powering our vacuum tube, which in turn powers a speaker, out will come a fierce current that varies exactly like the teeny signal current does. Loud music!
That takes us back to Bardeen and Brattain. The transistor they finally succeeded in making was a “point-contact” transistor, and despite its bizarre appearance, it’s really just two diodes, side by side, one with a small battery connected to it the “right” way and the other with a big battery connected to it the “wrong” way. The bigger circuit is switched on all the time, but because its battery is hooked up the “wrong” way, no current flows through it. However, the more juice we give the small current, the more the big current is able to flow. The small-current diode somehow breaks the spell that’s on the big-current diode, and now we can do exactly what a vacuum tube does: use a smaller signal current to control a bigger power current.
First I’ll show a diagram of what their invention really looked like schematically, but then a simpler one. Here it is in its real shape:
But here it is in a more matter-of-fact way:
What we’re trying to do here is use the knob on the teeny little input current to make sound come out of the speaker run by the beefy output current. That knob could represent a microphone, or radio receiver, or phonograph needle, or whatever. We don’t want the speaker blaring at full blast all the time, though; we only want it on when we say so, and at just the level we say. That’s amplification.
William Shockley (Bardeen and Brattain’s boss) used to say that if you want to understand amplification, think of lighting the hay under a mule’s butt on fire. It doesn’t take much energy for you to strike that match, but boy, do you get a lot of energy out the other end!
In the center of our pictured scheme above, we have a diode, made up of two types of semiconductor material, p-type and n-type, stuck together. Remember that a diode only allows current to run in one direction, and here we’ve set it up so that the input current can run, but the output current cannot. We say that in the input circuit, our diode has “forward bias”, which means current can run through it easily. But the diode in the output circuit has “reverse bias” because the battery is connected the other way around, and no current can flow through a diode like that.
One thing we need to know is: WHY does a diode made out of semiconductors only allow current to run in one direction?
We have to look a little closer at n-type and p-type semiconductors. I’m using silicon as an example because it’s the most common semiconductor material today.
On the top left you see a silicon atom. It has 4 electrons (negative charge, red) in its outer shell, which means they’re available to interact with other atoms. We’ll put 4 protons (positive charge, green) in the nucleus to balance out the charge and to remind ourselves that each atom’s overall charge is neutral. (A silicon atom really has 14 protons and 14 electrons, but I don’t think we want to draw all of those, so let’s keep it simpler here.)
We can grow a silicon crystal out of these atoms, and as you can see on the bottom left, all the electrons pair up (because they are happiest that way), and we can keep on adding more atoms to produce our crystal. Because the electrons really enjoy being paired up, they don’t move around very much.
Instead of growing a plain silicon crystal, though, we can throw in a dash of a different type of atom and get a different type of semiconductor material. Phosphorus atoms have 5 electrons in their outer shell, so if we put some of those in, there will be places in our finished crystal that have an extra electron sticking out (bottom center). It’s not paired up, so it moves more easily. That material is an n-type semiconductor (“n” for negative, to reflect the electron’s charge).
Or we could include some boron atoms, which only have 3 electrons in their outer shell. This crystal (bottom right) will have spaces in it that electrons can fill so they can pair up and be happy, and we call those spaces “holes”. Holes can’t actually move, but if an electron fills the hole, it leaves a hole somewhere else, so it’s as if holes can move. A semiconductor with extra holes in it is p-type, where “p” is for positive.
Even though these semiconductors are called “n-type” and “p-type”, we have to remember that they don’t have any overall electrical charge. They’re neutral to start with.
Now what happens when we put a p-type chunk next to an n-type chunk?
We’re not applying any voltage yet, but already some things are happening. Because there are a ton of free electrons (red) on the n side and almost none on the p side, electrons will start drifting over from n to p, just like a drop of dye will disperse in water. And there are holes (white) over there on the p side, so the electrons will fill them!
Both materials were electrically neutral to start with, but now they aren’t quite anymore. When the extra electron jumps off a phosphorus atom, it leaves behind a positive charge. Likewise, when an electron fills a hole in a boron atom, it brings along a negative charge.
So those electrons will keep drifting until the positive charge that’s building up on the n side becomes strong enough to counteract that drifting. When everything settles down, we’re left with a region in the middle with some separated charge, but also without any charge carriers that are good at moving around — just a lot of paired-up and cozy electrons. That middle region with no extra electrons and no extra holes we call the “depletion region”.
Now let’s apply voltage!
Electrons (red) want to move toward positive, and holes (white) want to move toward negative. With the forward-bias setup, it’s easy for them to do so. Electrons and holes get pushed together and cancel each other out, and they keep seeing more of each other, so current can keep flowing.
But with a reverse bias, electrons and holes get pulled apart, and the depletion region gets huge. It’s like pulling on a rubber band. The more you pull, the stronger the force that opposes you. Electrons would love to flow in from the negative end and reach the positive charge on the other side, but as they start doing that and filling some holes, the big negative charge up ahead of them in the depletion region not only repels them, but leaves them with no way to get across, because there aren’t any holes in there. By applying reverse bias, we’ve effectively snipped our wire in half, so there’s no current.
Now, if we go back to the point-contact transistor, you see we have these two diode situations side by side. The forward-biased diode is running off the small battery, and everything’s just peachy over there. But the big battery is trying to run the reverse-biased diode with no luck.
Bardeen and Brattain had been piddling around with different configurations and materials for months on end, because they didn’t completely understand what they were doing. But they did know that charge carriers (electrons and holes) will diffuse, or spread out, like smoke. Maybe the big current would run if a few carriers could find their way from the small current into its depletion region, so that its electrons could make it across. The depletion region in this case wasn’t very wide, because the p-type layer was really thin, basically only a crust that had formed on the top of the n-type layer. So it wouldn’t take all that much.
Bardeen got the idea to put the leads going into the p-type layer really close together, hoping that running the smaller current would let enough carriers diffuse across to kick on the bigger curent. He calculated that the leads shouldn’t be more than about 0.002 inches apart, which would allow carriers to get across the gap in about 1 microsecond.
And that explains the pointy plastic thing. Brattain made Bardeen’s idea work by putting a piece of gold foil across the bottom of the triangle to serve as both electrical leads, one going into each circuit, and he sliced that foil right at the point of the triangle with a razor blade to make a very thin slit. The two circuits would thus be kept separate, as they needed to be, but they’d be very close together. Would running the little circuit be enough to allow some stepping stones to slide over into the depletion region so that the big current would run on demand?
They hooked up the power sources to what must have been their umpteenth longshot gizmo. One of the two men (we don’t know which) spoke over the small circuit, and they listened for an amplified voice.
It worked the first time they tried it, on December 16, 1947. And with that, the modern world was born.
Brattain wrote in his notebook:
This circuit was actually spoken over and by switching the device in and out a distinct gain in speech level could be heard and seen on the scope presentation with no noticeable change in quality. By measurements at fixed frequency in it was determined that this power gain was the order of a factor of 18 or greater.
Bardeen said in his notebook that they got power gains up to 40, and the reason was that charge carriers had sidled over from the small circuit into the big circuit.
So, decades before you wound up dancing to it at somebody’s wedding, Bardeen and Brattain had quite literally implemented the Electric Slide!
A week later, they demonstrated their transistor amplifier at group meeting. If you’ve worked in academia or R&D in industry, you know that sometimes group meetings can be a little snooze-inducing. But after this one, the fifteen or so attendees walked out the door understanding that the world was going to be different. And even they didn’t grasp the full magnitude of the change.
By 1953, the world’s first transistorized hearing aid became available to the public:
And in 1954, the first transistor radio:
And you know the rest of the story: Every electronic device in your home and your workplace uses transistors, and lots of ‘em. Transistors are used not only as amplifiers, but as logic gates and switches in computers, phones, controllers, and all sorts of things.
As I said at the top, though, you probably still have a working vacuum tube in your house, and you’d be really mad if it went away, because there’d be no more quick access to toasty Hot Pockets or steamy Trader Joe’s frozen shu mai. That’s right … it’s in your microwave oven!
You still need to use the Dr. Frankenstein approach in there, because the electrons zipping through the space in a vacuum tube must zoom past little apertures in order to generate microwaves, sort of the way blowing into a flute generates sound waves. That particular type of vacuum tube is called a magnetron:
The boxy TV you finally replaced with a flat screen in the 1990s or 2000s was also a vacuum tube — a cathode ray tube — and that’s why the back of it needed to get hot, and that’s why it took a minute to “warm up” when it was turned on.
Some hardcore audiophiles still use vacuum-tube amplifiers for their stereos or guitars because they say the sound is better than with solid-state electronics. I’ll take their word for it.
You might also have a vacuum tube in your house if you operate a radar station or a broadcasting tower, but I’m going to guess most of you don’t do those things.
Transistor technology has certainly come a long way since Bardeen and Brattain’s funky coat-hanger-lookin’ doo-hickey, with the advent of bipolar junction transistors and especially field-effect transistors (now by far the most common kind). But if it weren’t for their dogged perseverance in formulating theory after theory and contraption after contraption, you wouldn’t be reading this right now.
And if you did, and you made it all the way to the end, I thank you so much for taking this trip through history with me.
Happy birthday, modern world! Hey, I hear there’s an open bar…