Cover of The Transistor, a book by Ketineni RamaraoFrom the book The Transistor
Chapter 31. Read the full book free as a PDF.

Two lineages in one device

Power switching had two options. A bipolar transistor carries large currents with a small voltage drop, but needs a steady drive current at its base. A power MOSFET is switched by voltage with almost no drive current, but loses more power in conduction at high voltage. The insulated gate bipolar transistor, the IGBT, fuses them: a MOS gate on the input side and a bipolar structure for the conduction. Easy control from the first, low conduction loss at high voltage from the second. For the two parent types see NPN versus PNP transistors and the FET transistor explained.

Close view inside an open power module: green base with rows of silicon chips in orange-bordered squares, linked by dozens of thin parallel bonding wires.
Inside an IGBT power module, a Mitsubishi CM600DU-24NFH with its cover removed. The square chips are the transistors and diodes, joined by rows of bonding wires that carry the current. Photo: Pslawinski, CC BY-SA 3.0, via Wikimedia Commons.

Made at General Electric

B. Jayant Baliga conceived the IGBT at General Electric's research centre in Schenectady, New York, around 1980. GE's chief executive, Jack Welch, wanted technology that would give the company's many businesses a lead, and he backed the project. In Baliga's own account, in January 1981 he flew to Cupertino, California, where Nathan Zommer ran a production line for power MOSFETs at the recently acquired Intersil. Zommer agreed to run IGBT wafers if Baliga supplied the material and the mask design. In the same account, the first wafers came off in August 1981 with 600 volt blocking capability and an on-state voltage drop of about 1 volt at 100 to 200 amps per square centimetre.

Baliga's account also dates the first IGBT product, GE's power-MOS "IGT" D94FQ4, to 20 June 1983, rated 400 and 500 volts at 18 amps, and a 5 horsepower adjustable-speed motor drive for heat pumps built on the new device to 3 October 1983. He says the company had held the technology back at first, but a paper submitted by a GE colleague in September 1982 made the idea public.

Where IGBTs are used

IEEE Spectrum lists household appliances, factory robots, lighting, cars, trains, televisions and photovoltaic installations: anywhere moderately high voltages must be switched on and off quickly. Baliga himself describes the IGBT as the power device of choice for medium and high power applications where operating voltages exceed 200 volts. It makes adjustable-speed motor drives efficient, because the drive can deliver only the power the motor needs instead of running flat out.

What is inside a module

A power module packs several IGBT chips and diodes onto one base, with rows of bonding wires to carry the current. The photograph above shows a Mitsubishi module with the cover off. The chips are the transistors, and the reason there are so many wires is that each wire can only carry a limited current. The module has to be cooled well, since the losses in conduction and switching all turn into heat.

Why it belongs in the transistor story

The IGBT shows that the transistor did not end with the microprocessor. A phone chip moves bits. An IGBT moves energy. Both come from the same three ideas of a gate, a channel or junction and a thin semiconductor structure. The thread starts in the history of the transistor.

Quick answers

What does IGBT stand for?

Insulated gate bipolar transistor. It has a MOS-style insulated gate for control and a bipolar structure for carrying current.

Who invented the IGBT?

B. Jayant Baliga, working at General Electric, conceived it around 1980. GE released the first IGBT product in 1983, according to Baliga's own account.

Where are IGBTs used?

In motor drives, appliances, factory robots, lighting, cars, trains and solar power equipment, where high voltages must be switched quickly and efficiently.

Go deeper in the book: The Transistor · PDF · Source Companion · Browse the category

Sources

The dates, the 600 V and 1 V figures and the D94FQ4 specifications are from Baliga's own first-person account, as in his June 2015 IEEE account and his IEEE Power Electronics Magazine and Wiley-IEEE writing, so they are his account and not independent confirmation. Uses are from IEEE Spectrum.

More from this book

History · From the book The Transistor

The history of the transistor

The transistor was not found by luck.

Explainer · From the book The Transistor

How transistors work

A transistor lets a small electrical signal control a much larger flow of current.

Explainer · From the book The Transistor

NPN vs PNP transistors: what is the difference?

An NPN transistor and a PNP transistor are mirror images.

Explainer · From the book The Transistor

The FET transistor and the MOSFET, explained

The field-effect transistor was patented in the 1920s and could not be built for another thirty years.