ABSTRACT
While the device just analyzed is called a coefficient multiplier when it is used in an analog computer, when used in an instrumentation context it would simply be an amplifier, in our example case an amplifier with a gain of 100 volts/volt. It might be used to amplify voltage from a strain-gage pressure sensor (which might supply O.IOO volt for full-scale pressure) to the IO-volt range for entry into an analog-to-digital converter, which requires a I 0-volt full-scale input signal. You might at this point be thinking, "Why don't you just use the op-amp directly for your amplifier since it has more than enough gain?" The problem here is that the op-amp gain can be relied upon to be "very large" but cannot be relied upon to be an accurate stable value. That is, the gain A is "guaranteed" only to be, say, in the range I to 5 million V /V. One op-amp that you might buy would have I.6 and another of the same type might have 3.8. Also, any one op-amp might have 1.6 on Tuesday and 2.8 on Wednesday. Such uncertainty in gain is totally intolerable in precision applications. Note, however, that as long as A is larger than the value used in a design study, the approximation used to get to Eq. (3-56) will be valid. Note that the gain of 100 that we achieved in our example depends for its accuracy and stability on the values of two fixed resistors, and not on the value of A, so long as A is "large enough." This is one of the main reasons why op-amps are so valuable as basic circuit building blocks. Using them, we can construct circuits whose performance depends mainly on passive components such as R and C, which can be selected to have accurate and stable values. The op-amp gain can "wander around" without causing any problem, so long as it always is "large enough."
