Field-Programmable Gate Arrays: Reconfigurable Logic for Rapid Prototyping and Implementation of Digital Systems

2.5   ELECTRICAL ASPECTS

While from many points of view we may regard FPGAs as consisting of idealized logic elements, in practice we need to be aware of the ways in which they are realized as electrical circuits, and their operation as analog circuits.

2.5.1   Complementary MOS Circuits

Almost all FPGAs use CMOS integrated-circuit technology. It uses two types of MOS transistor, referred to as N- and P-type. N-type transistors employ electrons as mobile charge carriers, while P-type transistors use holes instead. Figure 2–16 shows a complementary pair connected to form an inverter. Since we know that holes will move toward a lower potential and electrons toward a higher one, we can label the source (S) and drain (D) terminals as shown. The N-type transistor conducts if the voltage between its gate (G) and source exceeds a small positive threshold voltage. The P-transistor conducts if its gate-to-source voltage is more negative than a small (negative) threshold voltage. So if the input voltage to the inverter is zero (logic false), the P-transistor conducts, while the N-transistor does not. This causes the output voltage to be raised to the supply voltage Vdd, by charging the output capacitance. If the input voltage is high, corresponding to logic true, the N-transistor conducts, discharging the output capacitance and producing a logic false value. CMOS output voltages swing between the full range of the supply rail voltages, and can tolerate significant deterioration in input ...