Here at The Automation Engineer we find that one essential component in automation comes up so often in conversation that we were surprised that we hadn’t wrote anything about it yet; the humble IGBT.
Firstly, let’s define what we mean by an IGBT. It’s an acronym for insulated-gate bipolar transistor. Put simply, it’s an electronic switch. What makes the IGBT so special is that is highly efficient and fast. These are ideal features for electronically controlling speed. You’ll find IGBTs in variable-frequency drives (VFDs), electric cars, trains, variable speed refrigerators and air-conditioners.
You could say the IGBT is the critical element of future technology. But it’s taken many years to get to this point in our journey.
It all began in 1831.
Not yet the Victorian Era. And the first industrial revolution is beginning to take its grip.
At that time the very foundation of power electronics had yet to be laid. It starts with Michael Faraday. His work on the properties of induction lead to the breakthrough development of the power transformer in that year. Around this time scientists Nikola Tesla and Thomas Edison also made significant contributions.
It wasn’t until 1902, the beginning of the Edwardian era, when Peter Cooper Hewitt invented the mercury arc rectifier. At this point it was possible to convert alternating current power to direct current. It was used throughout industry. This technology triggered inventors and scientists to look for new means of power electronics, largely driven by the demands of radio and television.
Big change came with the successful testing of the bipolar point-contact transistor, by Brittain, Bardeen and Shockley at the Bell Telephone Laboratory. This is was the first type of solid-state electronic transistor capable of amplifying or switching electronic signals. It heralded the coming of a new era in power electronics.
Back then technology created limitations. To create the transistor required two metal contacts within 0.002 inches of each other. Wires back then were too thick. It was Brattain’s genius idea to attach a strip of gold foil over a triangle. He then sliced through the tip with a razorblade to create a hairline cut. The triangle was then suspended on a spring over a crystal of germanium (known for having high resistance). It was set so that the contacts lightly touched. The germanium crystal itself sat on a metal plate attached to a voltage source. This produced a small current through one contact, and a larger current through the other. In essence, amplifying the signal.
This works because germanium is a semiconductor. It can either let lots of current through or none at all. The germanium in the experiment had lots of electrons. As the electric signal travelled through into the gold foil it injected holes (these holes are like the opposite of electrons). The end result is a layer of germanium with too few electrons.
To explain this further, there’s a few terms to consider. When a semiconductor has too many electrons it’s known as N-Type. When there are too few, it’s a P-type. The place in-between is known as a P-N junction. This P-N junction is where current flows from one side up to the other.
The original design was unreliable and noisy. So in 1948 William Shockley developed the bipolar junction transistor or BJT to resolve the issues. The concept was a transistor that resembled a sandwich, with one type of semiconductor surrounding a second type. This new method was the result of careful testing which proved that electricity can travel through a semiconductor, not just around it. To do so though required a middle that was both very thin and pure.
To produce this sandwiched crystal Shockley worked alongside Gordon Teal. Teal believed there was a better way to create the sandwiched crystal, suggesting it should be grown not cut. His method used a tiny crystal and dipped it into melted germanium, pulling the larger crystal out as it solidified. Even though his peers were not interested, Teal continued his research. Later Shockley admitted he was wrong: grown crystals lasted 100 times longer than the cut ones.
Once on-board Shockley asked for impurities to be added to the melt to change the number of electrons (type N and P). By adding gallium they could turn the N-type germanium into a P-type, and back again by adding antimony. They improved the design by pulling the crystal even slower still and by stirring the melt. This new transistor type couldn’t handle the extremely rapid signal fluctuation, but they worked better in every other way.
By 1952 R.N Hall introduced the first power diode which had a larger PN junction than its predecessors. It was capable of handling 200V and had a current rating of 35A. Unfortunately by increasing the PN junction meant that power diodes were inappropriate for high voltage applications above 1MHz.
In 1956 technology changed once again, this time with the introduction of the Silicon Controlled Rectifier (SCR) by General Electric. This marked the transition from germanium to silicon as the key semiconductor material. The lower number of free electrons and higher operating temperature all contributed toward the change. In addition there’s an abundance of material (35% of the Earth’s surface is Silicon).
Technology leapt ahead again in the 1960s with MOSFET (metal-oxide-semiconductor field-effect transistor) which made it possible to control the load current with almost no input current. The metal-oxide aspect of the name is now a misnomer as the gate material is often silicon based. MOSFET made it possible for the development of digital circuits, often you’ll find thousands or millions within one single memory chip or microprocessor. They can be made as a p-type or n-type semiconductor, and work as complimentary pairs of MOS transistors, known as CMOS.
Throughout the 1970s the US and Japan were the powerhouse of the FETs and the semiconductor industry. In 1969 the Japanese Electrotechnical Laboratory invented the V-groove MOSFET, and in 1976 power MOSFET became commercially available.
Then in 1982, a cheap, robust and fast device was developed with the capability to turn off and on rapidly, called the Insulated Gate Bipolar Transistor (IGBT). It had taken a number of years to develop this specific technology, but it started with a researcher called Yamagami who filed his first patent in 1968. His proposal involved a MOS controlling a positive-negative-positive-negative (PNPN) semiconductor without using any of the output voltage.
Later, the two largest electronics companies of the era developed the IGBT to its potential. To this day both Baliga of GE, and Becke and Wheatley of RCA contest the other as the originator. The reality is both had their patents approved which accelerated the process.
One of the most common uses for the IGBT is within the inverter stage of a variable speed drive. To understand what the IGBT does within the drive, we need to understand what the IGBT does in general. Imagine it as a gate. When voltage is below a certain level, the gate won’t open. But when the voltage is above a set amount, it allows the current to move. It can perform this action within nanoseconds.
Inside the AC drive, acting as a switch, the IGBT controls the amount of voltage going to the motor using Pulse-Width Modulation (PWM). This then allows the user to control the speed of that motor. There are usually six IGBTs in a drive which work together to replicate a waveform. The longer the cycle, the slower the speed.
Today it is clear that without the development of IGBTs our world would be very different indeed. From the early pioneers like Faraday, Tesla and Edison, to innovators like Hewett, Shockley and Teal, and the modern power electronics of today, each one paved the way for future technology. It’s this technology that will make our world habitable for future generations.