Skip to content
LearnTronics menu

LearnTronics

Electronic Parts

Discrete Components

Transistors

From Transistors to Integrated Circuits

Once transistors could be made reliably in silicon, the next question was obvious: why manufacture each transistor as a separate component, mount every one in its own package, and wire them together by hand?

The problem of many parts

A complex transistor circuit might require hundreds or thousands of discrete transistors, diodes, resistors, capacitors, sockets, wires, and solder joints. As circuit complexity grew, the physical connections became a major source of size, cost, weight, power, and failure.

Put the circuit in the semiconductor

The integrated-circuit idea is to fabricate many circuit elements in and on the same piece of semiconductor and connect them with patterned metal layers. Instead of wiring together packaged transistors, lithography creates the transistors and their interconnections together.

The planar process

Jean Hoerni's planar process at Fairchild in 1959 left protective silicon dioxide over the wafer surface and junctions. The oxide could be opened in selected places for contacts while protecting the rest of the device. This was a crucial manufacturing step toward reliable monolithic integrated circuits.

Kilby and Noyce

Jack Kilby at Texas Instruments demonstrated an integrated circuit in 1958 using semiconductor components formed on one piece of germanium and connected with fine wires. Robert Noyce at Fairchild soon proposed a monolithic silicon approach using Hoerni's planar process and deposited metal interconnections, which was much better suited to mass production.

Why MOSFETs became dominant

MOS transistors are especially well suited to dense integrated circuits because their insulated gates require little DC current and the structures can be scaled to very small dimensions. Complementary MOS—CMOS—uses N-channel and P-channel MOSFETs together to produce logic with very low static power.

A logic inverter

The simplest CMOS logic gate uses one P-channel MOSFET and one N-channel MOSFET. When the input is low, the P-channel device pulls the output high. When the input is high, the N-channel device pulls the output low.

That two-transistor circuit is the seed from which enormous digital systems can be built.

From gates to processors

Logic gates combine into flip-flops, counters, adders, registers, memories, state machines, and processors. Analog circuits similarly combine matched transistors into differential amplifiers, current mirrors, references, oscillators, regulators, and data converters.

Discrete transistors still matter

Integrated circuits did not make discrete transistors obsolete. A tiny IC may control a separate power MOSFET that switches tens or hundreds of amperes, a low-noise JFET may sit at the input of an instrument, and a discrete RF transistor may handle frequencies or power levels unsuitable for a general integrated process.

The transistor therefore sits at a useful dividing point in LearnTronics: it is both a discrete component that can be held in your hand and the microscopic building block from which integrated electronics is constructed.
Next study

From here the natural progression is into integrated circuits: first simple logic gates and operational amplifiers, then increasingly complex digital and analog systems. The important idea remains the same—large electronic functions are created by combining many simple transistor actions in carefully arranged circuits.