CMOS logic technology

Complementary metal-oxide-semiconductor logic, almost always shortened to CMOS, builds each gate from pairs of field-effect transistors: a p-channel device that conducts to the positive supply and an n-channel device that conducts to ground. The two are arranged so that for any steady input state one transistor of each pair is switched off, which means the gate draws virtually no current except while it is changing state. This is the key reason CMOS dominates modern integrated circuits, from microprocessors to memory.

Because the power a CMOS circuit consumes is mostly dynamic, it rises with switching frequency and with the supply voltage, but at rest the static power is extremely low. CMOS also offers high noise immunity, since its switching threshold sits near the middle of the supply range, and its inputs present a very high impedance because they connect to insulated transistor gates. A widely used early standard was the 4000 series of CMOS chips, which could run over a broad range of supply voltages, a flexibility that made it popular for general-purpose and battery-powered designs.

The high gate impedance brings one practical concern: the thin insulating layer can be damaged by static electricity, so CMOS parts include protection diodes and should be handled with care. Unused inputs must be tied to a defined level rather than left floating, because a floating gate can drift to an intermediate voltage and cause both transistors to conduct. Over successive generations CMOS feature sizes have shrunk dramatically, raising speed and density while lowering operating voltage.

Frequently asked questions

Why does CMOS consume little power at rest?

In any stable state one transistor in each complementary pair is off, so almost no current flows between the supply rails except briefly during switching.

Why must unused CMOS inputs not be left floating?

A floating high-impedance input can drift to a mid-level voltage that turns both transistors partly on, wasting power and producing unreliable output, so spare inputs are tied high or low.

What was the 4000 series?

It was an early and widely used family of CMOS logic chips known for operating across a wide supply-voltage range, which suited many low-power and general designs.