Why is the Triode operating region of a MOSFET named such?

If triodes were used for amplifiers, why is the triode region generally used for digital and the saturation region generally used for analog? Am I missing a fundamental understanding of the MOSFET's characteristics?

• Triodes were not only used as amplifiers but also as switches. Early computing machines were made up of relays and then tubes. The vacuum tube based Eniac was a digital machine. – Kaz Oct 20 '12 at 22:01

Why is the Triode operating region of a MOSFET named such?

When a FET is operating in the triode (ohmic) region, the drain current is strongly dependent on drain-source voltage. The plate current of a vacuum tube triode is also strongly dependent on plate-cathode voltage and this is, I believe, why the triode region of operation is so named.

FET drain curves operating in triode region:

Vacuum tube triode plate curves:

This strong dependence of drain (plate) current on $$\v_{DS}\$$ ($$\v_{PK}\$$) severely limits the available voltage gain of the device. The greater the dependence, the less available gain as this dependence is, in effect, acting as a resistor in parallel with the load, shunting signal current around the load.

So, although triodes were (and still are) used as amplifiers, tetrodes and pentodes were developed that, like the saturation region for FETs, greatly reduce the plate current dependence on plate-cathode voltage. These devices offer, among other things, much higher amplification factors, i.e., they have much higher plate impedances.

The amount of current which passes through a MOSFET will be a function of the voltage on the gate and drain. A given gate voltage, at low drain-source voltages, the drain-source current will be roughly proportional to the drain-source voltage ("triode mode"). At higher drain-source voltages, the current will be essentially constant ("saturation mode"). At intermediate voltages, the behavior will be somewhere between those two "modes".

Amplfiiers often use transistors in saturation mode because, provided there is a wide range of drain voltages where Ids is essentially independent of Vds. By contrast, there is a comparatively narrow range of drain voltages where Ids is reasonably proportional to Vds. Further, it's easier to design circuit elements to behave in reasonably-linear fashion with Ids is independent of Vds, than with Ids proportional to Vds. People designing vacuum-tube amplifiers might have preferred to use their amplifying elements in a fashion where anode current was independent of anode voltage, but vacuum tubes don't work that way. Triodes were used in "triode mode" because, well, they were triodes.

As for digital circuits, the normal goal is to turn on MOSFETs "as hard as possible". If the amount of sustained current through a MOSFET would be large enough to put it into saturation, that's usually a sign that either the MOSFET should be bigger, or the amount of current drawn through it should be smaller.