AN-849

Rev. A | Page 2 of 8

Most comparators are quite fast and some are very fast indeed—but

so are some op amps. Why must we expect low speed when using

an op amp as a comparator? A comparator is designed to be used

with large differential input voltages, whereas op amps typically

operate closed loop with their input voltages minimized by negative

feedback. When an op amp is overdriven, sometimes by only a few

millivolts, some of its stages may saturate. If this occurs, the device

takes a comparatively long time to come out of saturation, and is,

therefore, much slower than if it were unsaturated.

OUTPUT OF OP AMP USED

AS A COMPARATOR

OUTPUT OF OP AMP USED

CLOSED-LOOP

(NO SATURATION/

DESATURATION)

INPUT

DELAY DUE TO

DESATURATION

t

0

Figure 2.

The desaturation time of an overdriven op amp is likely to be

considerably longer than its normal group delay (effectively the

time a signal takes from input to output) and often depends on

the amount of overdrive.

Because few op amps are characterized for their recovery times

from differing levels of overdrive, it is necessary for the user to

determine, by experiment, the delays likely to result from the

levels of overdrive expected in a particular application. The

value used in design calculations should be at least twice the

worst value seen in any tests, because the samples tested may

not be representative.

The output of a purpose-built comparator is designed to drive a

particular logic family or families. Often there is a separate supply

for the output stage to ensure that the logic levels are correct.

A modern op amp usually has rail-to-rail output—its maximum

positive level is close to the positive supply, and the minimum

negative level is close to the negative supply. (Older designs used

structures with headroom of more than 1.5 V to both supplies.)

OP AMP

LOGIC

Figure 3.

If logic and op amp share the same supply, a rail-to-rail op amp

can drive CMOS and TTL logic families successfully, but if the

op amp and logic have different supplies, additional interface

circuitry is necessary between them. Note that this applies to an

op amp with ±5 V supplies that must drive logic with +5 V

supply; the logic is liable to be damaged if −5 V is applied to it.

The simplest interface circuits are inverters. They may be built

with NPN transistors, but these draw base current. N-channel

MOSFET transistors are more convenient.

OP AMP

LOGIC

LOGIC

OP AMP

Figure 4.

current. The lower their value, the faster the inverter, but the

more power it consumes. Values of a few kilohms are usually

used. The N-channel MOS device should have a low gate

threshold voltage (<2 V) and a gate-source breakdown voltage

greater than the maximum output voltages of the op amp.

Usually, ±25 V is sufficient. Even though the op amp and the

logic use different supplies, they must be related to each other.

voltage to turn on the transistor or the MOSFET gate. In

addition, the op amp negative supply must not be positive

relative to the logic negative supply, although it may be

connected to it. One must, of course, observe the absolute

maximum ratings of all the devices used.

LOGIC

OP AMP

P

N

Figure 5.

A complementary MOS inverter can be built with a P-channel

and an N-channel MOSFET. This has the advantage of drawing

no standby current, but it does have a large current spike during

switching when both devices are briefly turned on together.

addition, the op amp negative supply must be equal to or less

than the logic negative supply.