I read in a book three opamp instrumentation amplifiers are better and they are more common, but why is it better than the single opamp instrumentation amplifier? It must be more expensive as they have two extra opamps.
First, many InAmps are integrated into one IC, also the three-opamp versions. Then the cost of the extra opamps isn't that high. Consider the extra cost of less than a mm^2 real estate against the total cost to create an IC.
Next, I've never seen a single-opamp inamp. There are two-opamp and three-opamp versions, and there's a reason why you can't make one with a single opamp: you need high impedance for both inputs, and that means neither can have feedback.
You probably mean a differential amplifier, like The Resistance comments below. Notice that the 2 input opamps are configured as non-inverting amplifiers, and that there's no feedback to the input pins, like you would have for an inverting amplifier. This way you get high-impedance amplifying buffer-amps before the actual differential amplifier.
There are a number of different issues in play here. Assuming you're talking about a three-op-amp instrumentation amplifier and a one-op-amp differential amplifier.
First, lets talk about hand-made instrumentation-amps and differential amplifiers vs. purpose-built integrated circuits. Hand-made amplifiers of both classes will not be as good as the IC's, which have laser-trimmed resistors providing very high common mode rejection ratios (CMRRs), and are often a good deal worse, even with 1% tolerance resistors.
That out of the way, while we're on CMRR, both in-amps and differential amp IC's have very high CMRR, which is largely determined by how well the resistors in the output stage are matched (plus how well some resistors on the input stage are matched for the op-amp). Since in-amps have an additional gain stage, and are often used for high gain circuits than diff-amps, the in-amps often have better CMRR, but it would be about the same as a diff-amp if the gain of the input stage was unity.
Both in amps and diff amps take a differential input and convert to a single ended output. The in-amps have the distinct advantage that you can change the gain with one resistor, and doing so does not impact CMRR. Diff amps usually have one gain.
The in-amps also have the advantage of providing both high and equal input impedances to both differential inputs. Diff amps provide lower input impedance, and the input resistance is not equal bot both inputs.
One advantage of the differential amps is that because the input to the actual amplifier goes through a voltage divider, the range of input voltages that the chip will work for at a given Vss/Vdd tends to be larger than that of an in-amp (in fact, double for a differential amp with a gain of 1). Other than that, and possibly cost, I can't think of many advantages the diff amp has over the instrumentation amp.
IA's are not just more expensive because of more op amps, rather the process includes laser trim on resistor values so offset voltages and fixed gain options offer unbeatable performance.
Some also add extra features that must be integrated for extreme high temperature, such as low offset, low noise and 126 dB CMRR with 40db gain and drift of 3 nV/V/°C with operation to +210°C in extreme high quality ceramic.
A standard one op-amp differential amplifier (which is what I believe you meant) is a poor comparison to the IAs you get from various sources. Here are a few things that make the one op-amp version much less attractive when measuring small signals: -
- They need four resistors to work. (An IA only normally uses one)
- The four resistors have to be balanced to avoid poor common mode problems. For an IA, the single resistor has only to be "accurate" to be able to predict gain performance.
- The input impedance is definitely finite and this can cause problems on some measurements. An IA uses only "floating" inputs (within the constraints of the power rails.
- The input impedance of the "negative" input is different to the "positive" input i.e. earth impedance balance is poor.
Hope this helps.