|
Every time Intel or AMD releases a new processor, the technogeeks amongst us rave on how many transistors they had crammed into it. The number used to be in the thousands, then millions, and now billions. But what are these things that we associated with power, speed and better gaming experience?
These humble devices have a unique history, and without it much of our civilisation would not have existed. |
"What's not fully realized is that Moore's Law was not the first paradigm to bring exponential growth to computers. We had electromechanical calculators, relay-based computers, vacuum tubes, and transistors. Every time one paradigm ran out of steam, another took over."
Ray Kurzweil, 'The New Humanists: Science at the Edge', 2003
Many inventions are conceived simultaneously by several different persons because the time is "right", meaning that a technical and scientific foundation exists and that there is demand and business potential for the invention.
The transistor, however, is an invention that was conceived long before the time was right. It was invented in 1947, and even several years later, it was considered by a scientific conference to be such an odd accomplishment that it was not included in the documentation. The inventors themselves believed that the transistor might be used in some special instruments and possible in military radio equipment. Yet the transistor is fundamental for all modern technology, including telecommunications, data communications, aviation and audio and video recording equipment.
Three persons, William Shockley, John Bardeen and Walter Brattain, shared the Nobel Prize in Physics for the breakthrough that they achieved on December 23, 1947. In certain respects, a fourth person was responsible for the transistor being discovered at that time. Marvin Kelley, who was then head of Bell Laboratories, had brought the trio together. Kelley believed that working with such an unknown group of materials as semiconductors demanded a combination of different specialties: the brilliant theorist Brattain, the skilled materials expert Bardeen and the very accomplished experimentalist Shockley, who was also strong on theory. The objectives for the project were very general.
The transistor, however, is an invention that was conceived long before the time was right. It was invented in 1947, and even several years later, it was considered by a scientific conference to be such an odd accomplishment that it was not included in the documentation. The inventors themselves believed that the transistor might be used in some special instruments and possible in military radio equipment. Yet the transistor is fundamental for all modern technology, including telecommunications, data communications, aviation and audio and video recording equipment.
Three persons, William Shockley, John Bardeen and Walter Brattain, shared the Nobel Prize in Physics for the breakthrough that they achieved on December 23, 1947. In certain respects, a fourth person was responsible for the transistor being discovered at that time. Marvin Kelley, who was then head of Bell Laboratories, had brought the trio together. Kelley believed that working with such an unknown group of materials as semiconductors demanded a combination of different specialties: the brilliant theorist Brattain, the skilled materials expert Bardeen and the very accomplished experimentalist Shockley, who was also strong on theory. The objectives for the project were very general.
Shockley
|
Bardeen
|
Brattain
|
Bell Laboratories in the US was part of one of the world's leading telephone companies, AT&T. The company realized that the transistor could be used for applications far removed from telecommunications in the strictest sense and decided, perhaps to avoid being accused of exploiting a monopoly position in its domestic market, to offer licenses on reasonable terms to all companies who wished to apply. In exchange, these companies were asked to contribute their own patents to a common patent pool.
In computers, as in radio and TV equipment, electronic tubes were used that were relatively bulky and consumed considerable energy. Designers, however, knew how to make them smaller, and factories knew how to make them reliably and at low cost. The new transistors, on the other hand, were fragile, could not withstand high temperatures and required much more complicated equations in design work. Telephone stations didn’t even use tubes. They were extremely reliable wonders of mechanical engineering based on relays and connecting rods.
Shortly before Brittain, Bardeen and Shockley were awarded the Nobel Prize, the first major application of the transistor had emerged. This was a small portable radio that was even called a transistor after the component that made it possible. Texas Instruments, which was the first company to introduce a radio of this type, would eventually achieve fame in the new semiconductor industry. The second company, which would become a giant in the consumer electronics industry, was Japanese. That company, which was started after World War II, had international ambitions and thus chose the English-sounding name Sony.
William Shockley had not been present on the day that the transistor worked for the first time. In his anger, at least according to the legends, he then sat down and invented a number of different varieties of transistors. These were based on how the transistor's three contacts were created by soldering, using diffusion under heat, etc. All of these variants are based on the method used to create the different layers through which current is controlled by a signal to an electrode in the middle of the three. Less than a decade later, another principle called the field effect was developed in which the size of the channel through which the current flows is controlled. A Swede named J Torkel Wallmark, who at that time worked at RCA in the US, played a key role in this invention.
What is a Transistor Anyway?
A transistor is really simple—and really complex. Let's start with the simple part. A transistor is a miniature electronic component that can do two different jobs. It can work either as an amplifier or a switch.
When it works as an amplifier, it takes in a tiny electric current at one end (an input current) and produces a much bigger electric current (an output current) at the other. In other words, it's a kind of current booster. That comes in really useful in things like hearing aids, one of the first things people used transistors for. A hearing aid has a tiny microphone in it that picks up sounds from the world around you and turns them into fluctuating electric currents. These are fed into a transistor that boosts them and powers a tiny loudspeaker, so you hear a much louder version of the sounds around you. William Shockley, one of the inventors of the transistor, once explained transistor-amplifiers to a student in a more humorous way: "If you take a bale of hay and tie it to the tail of a mule and then strike a match and set the bale of hay on fire, and if you then compare the energy expended shortly thereafter by the mule with the energy expended by yourself in the striking of the match, you will understand the concept of amplification."
Transistors can also work as switches. A tiny electric current flowing through one part of a transistor can make a much bigger current flow through another part of it. In other words, the small current switches on the larger one. This is essentially how all computer chips work. For example, a memory chip contains hundreds of millions or even billions of transistors, each of which can be switched on or off individually. Since each transistor can be in two distinct states, it can store two different numbers, zero and one. With billions of transistors, a chip can store billions of zeros and ones, and almost as many ordinary numbers and letters (or characters, as we call them). More about this in a moment.
The great thing about old-style machines was that you could take them apart to figure out how they worked. It was never too hard, with a bit of pushing and poking, to discover which bit did what and how one thing led to another. But electronics is entirely different. It's all about using electrons to control electricity. An electron is a minute particle inside an atom. It's so infinitesimally small, it weighs just under 1x10-24 gram! The most advanced transistors work by controlling the movements of individual electrons, so you can imagine just how small they are. In a modern computer chip, the size of a fingernail, you'll probably find between 500 million and two billion separate transistors. There's no chance of taking a transistor apart to find out how it works, so we have to understand it with theory and imagination instead. First off, it helps if we know what a transistor is made from.
How is it Made?
Transistors are made from silicon, a chemical element found in sand, which does not normally conduct electricity (it doesn't allow electrons to flow through it easily). Silicon is a semiconductor, which means it's neither really a conductor (something like a metal that lets electricity flow) nor an insulator (something like plastic that stops electricity flowing). If we treat silicon with impurities (a process known as doping), we can make it behave in a different way. If we dope silicon with the chemical elements arsenic, phosphorus, or antimony, the silicon gains some extra "free" electrons—ones that can carry an electric current—so electrons will flow out of it more naturally. Because electrons have a negative charge, silicon treated this way is called n-type (negative type). We can also dope silicon with other impurities such as boron, gallium, and aluminum. Silicon treated this way has fewer of those "free" electrons, so the electrons in nearby materials will tend to flow into it. We call this sort of silicon p-type (positive type).
Quickly, in passing, it's important to note that neither n-type or p-type silicon actually has a charge in itself: both are electrically neutral. It's true that n-type silicon has extra "free" electrons that increase its conductivity, while p-type silicon has fewer of those free electrons, which helps to increase its conductivity in the opposite way. In each case, the extra conductivity comes from having added neutral (uncharged) atoms of impurities to silicon that was neutral to start with—and we can't create electrical charges out of thin air! A more detailed explanation would need me to introduce an idea called band theory, which is a little bit beyond the scope of this article. All we need to remember is that "extra electrons" means extra free electrons—ones that can freely move about and help to carry an electric current.
In computers, as in radio and TV equipment, electronic tubes were used that were relatively bulky and consumed considerable energy. Designers, however, knew how to make them smaller, and factories knew how to make them reliably and at low cost. The new transistors, on the other hand, were fragile, could not withstand high temperatures and required much more complicated equations in design work. Telephone stations didn’t even use tubes. They were extremely reliable wonders of mechanical engineering based on relays and connecting rods.
Shortly before Brittain, Bardeen and Shockley were awarded the Nobel Prize, the first major application of the transistor had emerged. This was a small portable radio that was even called a transistor after the component that made it possible. Texas Instruments, which was the first company to introduce a radio of this type, would eventually achieve fame in the new semiconductor industry. The second company, which would become a giant in the consumer electronics industry, was Japanese. That company, which was started after World War II, had international ambitions and thus chose the English-sounding name Sony.
William Shockley had not been present on the day that the transistor worked for the first time. In his anger, at least according to the legends, he then sat down and invented a number of different varieties of transistors. These were based on how the transistor's three contacts were created by soldering, using diffusion under heat, etc. All of these variants are based on the method used to create the different layers through which current is controlled by a signal to an electrode in the middle of the three. Less than a decade later, another principle called the field effect was developed in which the size of the channel through which the current flows is controlled. A Swede named J Torkel Wallmark, who at that time worked at RCA in the US, played a key role in this invention.
What is a Transistor Anyway?
A transistor is really simple—and really complex. Let's start with the simple part. A transistor is a miniature electronic component that can do two different jobs. It can work either as an amplifier or a switch.
When it works as an amplifier, it takes in a tiny electric current at one end (an input current) and produces a much bigger electric current (an output current) at the other. In other words, it's a kind of current booster. That comes in really useful in things like hearing aids, one of the first things people used transistors for. A hearing aid has a tiny microphone in it that picks up sounds from the world around you and turns them into fluctuating electric currents. These are fed into a transistor that boosts them and powers a tiny loudspeaker, so you hear a much louder version of the sounds around you. William Shockley, one of the inventors of the transistor, once explained transistor-amplifiers to a student in a more humorous way: "If you take a bale of hay and tie it to the tail of a mule and then strike a match and set the bale of hay on fire, and if you then compare the energy expended shortly thereafter by the mule with the energy expended by yourself in the striking of the match, you will understand the concept of amplification."
Transistors can also work as switches. A tiny electric current flowing through one part of a transistor can make a much bigger current flow through another part of it. In other words, the small current switches on the larger one. This is essentially how all computer chips work. For example, a memory chip contains hundreds of millions or even billions of transistors, each of which can be switched on or off individually. Since each transistor can be in two distinct states, it can store two different numbers, zero and one. With billions of transistors, a chip can store billions of zeros and ones, and almost as many ordinary numbers and letters (or characters, as we call them). More about this in a moment.
The great thing about old-style machines was that you could take them apart to figure out how they worked. It was never too hard, with a bit of pushing and poking, to discover which bit did what and how one thing led to another. But electronics is entirely different. It's all about using electrons to control electricity. An electron is a minute particle inside an atom. It's so infinitesimally small, it weighs just under 1x10-24 gram! The most advanced transistors work by controlling the movements of individual electrons, so you can imagine just how small they are. In a modern computer chip, the size of a fingernail, you'll probably find between 500 million and two billion separate transistors. There's no chance of taking a transistor apart to find out how it works, so we have to understand it with theory and imagination instead. First off, it helps if we know what a transistor is made from.
How is it Made?
Transistors are made from silicon, a chemical element found in sand, which does not normally conduct electricity (it doesn't allow electrons to flow through it easily). Silicon is a semiconductor, which means it's neither really a conductor (something like a metal that lets electricity flow) nor an insulator (something like plastic that stops electricity flowing). If we treat silicon with impurities (a process known as doping), we can make it behave in a different way. If we dope silicon with the chemical elements arsenic, phosphorus, or antimony, the silicon gains some extra "free" electrons—ones that can carry an electric current—so electrons will flow out of it more naturally. Because electrons have a negative charge, silicon treated this way is called n-type (negative type). We can also dope silicon with other impurities such as boron, gallium, and aluminum. Silicon treated this way has fewer of those "free" electrons, so the electrons in nearby materials will tend to flow into it. We call this sort of silicon p-type (positive type).
Quickly, in passing, it's important to note that neither n-type or p-type silicon actually has a charge in itself: both are electrically neutral. It's true that n-type silicon has extra "free" electrons that increase its conductivity, while p-type silicon has fewer of those free electrons, which helps to increase its conductivity in the opposite way. In each case, the extra conductivity comes from having added neutral (uncharged) atoms of impurities to silicon that was neutral to start with—and we can't create electrical charges out of thin air! A more detailed explanation would need me to introduce an idea called band theory, which is a little bit beyond the scope of this article. All we need to remember is that "extra electrons" means extra free electrons—ones that can freely move about and help to carry an electric current.
How do transistors work in calculators and computers?
In practice, you don't need to know any of this stuff about electrons and holes unless you're going to design computer chips for a living! All you need to know is that a transistor works like an amplifier or a switch, using a small current to switch on a larger one. But there's one other thing worth knowing: how does all this help computers store information and make decisions?
We can put a few transistor switches together to make something called a logic gate, which compares several input currents and gives a different output as a result. Logic gates let computers make very simple decisions using a mathematical technique called Boolean algebra. Your brain makes decisions the same way. For example, using "inputs" (things you know) about the weather and what you have in your hallway, you can make a decision like this: "If it's raining AND I have an umbrella, I will go to the shops". That's an example of Boolean algebra using what's called an AND "operator" (the word operator is just a bit of mathematical jargon to make things seem more complicated than they really are). You can make similar decisions with other operators. "If it's windy OR it's snowing, then I will put on a coat" is an example of using an OR operator. Or how about "If it's raining AND I have an umbrella OR I have a coat then it's okay to go out". Using AND, OR, and other operators called NOR, XOR, NOT, and NAND, computers can add up or compare binary numbers. That idea is the foundation stone of computer programs: the logical series of instructions that make computers do things.
Normally, a junction transistor is "off" when there is no base current and switches to "on" when the base current flows. That means it takes an electric current to switch the transistor on or off. But transistors like this can be hooked up with logic gates so their output connections feed back into their inputs. The transistor then stays on even when the base current is removed. Each time a new base current flows, the transistor "flips" on or off. It remains in one of those stable states (either on or off) until another current comes along and flips it the other way. This kind of arrangement is known as a flip-flop and it turns a transistor into a simple memory device that stores a zero (when it's off) or a one (when it's on). Flip-flops are the basic technology behind computer memory chips.
In practice, you don't need to know any of this stuff about electrons and holes unless you're going to design computer chips for a living! All you need to know is that a transistor works like an amplifier or a switch, using a small current to switch on a larger one. But there's one other thing worth knowing: how does all this help computers store information and make decisions?
We can put a few transistor switches together to make something called a logic gate, which compares several input currents and gives a different output as a result. Logic gates let computers make very simple decisions using a mathematical technique called Boolean algebra. Your brain makes decisions the same way. For example, using "inputs" (things you know) about the weather and what you have in your hallway, you can make a decision like this: "If it's raining AND I have an umbrella, I will go to the shops". That's an example of Boolean algebra using what's called an AND "operator" (the word operator is just a bit of mathematical jargon to make things seem more complicated than they really are). You can make similar decisions with other operators. "If it's windy OR it's snowing, then I will put on a coat" is an example of using an OR operator. Or how about "If it's raining AND I have an umbrella OR I have a coat then it's okay to go out". Using AND, OR, and other operators called NOR, XOR, NOT, and NAND, computers can add up or compare binary numbers. That idea is the foundation stone of computer programs: the logical series of instructions that make computers do things.
Normally, a junction transistor is "off" when there is no base current and switches to "on" when the base current flows. That means it takes an electric current to switch the transistor on or off. But transistors like this can be hooked up with logic gates so their output connections feed back into their inputs. The transistor then stays on even when the base current is removed. Each time a new base current flows, the transistor "flips" on or off. It remains in one of those stable states (either on or off) until another current comes along and flips it the other way. This kind of arrangement is known as a flip-flop and it turns a transistor into a simple memory device that stores a zero (when it's off) or a one (when it's on). Flip-flops are the basic technology behind computer memory chips.
Ponder this
What makes semiconductors to only semi-conduct electricity?
Why was silicon chosen as the ideal semiconductor for the first transistor? What advantages does it have versus other semiconducting materials?
Discuss
Discuss on how logic gates are constructed using transistors. How would the circuit look like? What kind of transistors (e.g. NPN, PNP) are used, and why? Design a logic gate that can solve simple mathematical problems such as long divisions and multiplications.
Further readings
Transistor, at Wikipedia
William Shockley, John Bardeen and Walter Brattain, the godfathers of the electronic age, at Nobelprize.Org
The Lost History of the Transistor, the story of how Texas Instruments and Bell Labs changed the world, at IEEE Spectrum
Carbon nanotube transistors, the next step? At SceinceAlert
