# Why is a Instrumentation Amplifier Necessary for A Wheatstone Bridge (small signal circuits)

I do not understand why an Instrumentation Amplifier is Necessary for small differential signals?

Why is it necessary that the initial inputs to the amplifier must be high impedance from the Wheatstone bridge? Since technically after the input voltages have been buffered and amplified, arent we back at the same differential amplifier stage that contains a low impedance input. Therefore wouldnt this low impedance input for the diff amp change the accuracy for the output?

My Questions

1. Are the input stage gain buffers required so that the wheatstone bridge voltages do not change? Since if it went into a differential amplifier immediately, the amplifier would draw current by itself due to its low impedance correct and so this may cause voltage inaccuracies?

2. After the buffers have outputted there voltage, does the differential amplifier not change the voltages since the low impedance can cause voltage drops etc? Or is this not the case since after the buffer, the output voltages (before R2) is like a source and so the amplifier draws the current separately. When compared to if just a differential amplifier was used the current would be the Wheatstone bridges resistance + the amplifiers resistance. Hence what I am trying to say is since the buffer separates the diff amp from the Wheatstone bridge the input voltage to the diff amp will be accurate.

1. Will there still not be a voltage drop across R2 and R3 which will change the input voltages to the differential amplifier from the output of the buffer gain amplifiers. Or is this ok for the differential voltage since these voltage drops will be constant and so the change in the inputs stays the same?

2. If number 3 is the case, then why is the low input impedance of a single differential amplifier bad if a Wheatstone bridge was directly connected to it? Since as these input impedances are constant wouldn't the voltage change still stay the same?

• It's not always necessary, but it does allow you to ignore some complications like the output impedance of the bridge. Aug 7, 2019 at 12:35

1. Yes, the input stage buffers are required so that there is a very high impedance load on the bridge and that the buffer load has virtually no effect on the resistors in the bridge.
2. The outputs of the buffers are very low impedance (typically less than 0.1 ohms at DC) and so the loading effect of the differential amplifier (inside the InAmp) has negligible effect.
3. There is a voltage drop across R2 and R3 because that is how a differential amplifier works but those resistors do not load the output from the buffers and cause an error.
4. Connecting a diff amp directly to a Wheatsone bridge will always cause errors that are circumvented by using buffers.
• Thanks, Andy, Just for number 3 so is the voltage drop from the resistors in the InAmp alright as this is a constant drop and so the change in voltage will still be the same as what the buffer amp output voltage was? Aug 7, 2019 at 9:05
• (3) You need to study how a diff amp works - the voltage dropped across R2 and R3 is zero when the common mode voltage equals the ground voltage at the bottom of R3. For all other scenarios there is a voltage dropped across R2 and R3. It is what it is and you should look how diff amps work. Aug 7, 2019 at 9:12
• (4) Loading the bridge is OK when the bridge is fully balanced. If not fully balanced then you will get a loading error term that also makes the bridge more non-linear. Aug 7, 2019 at 9:13
• I don't have any links that spring to mind. I used to design bridge front-end amplifiers so this knowledge is ingrained. InAmps are unambiguously easy to use but sometimes I've used op-amp diff amps if the loading effect was very minor and I needed both rail-to-rail AND low offset performance. An op-amp diff amp using high impedances will be good enough in many examples but, always be aware that resistor value matching is very, very important in either InAmp or diff amp. Aug 7, 2019 at 9:37
• I'm referring to mismatches in value and temperature coefficients. For instance, the ratio of R2 to R3 has to be matched in both limbs of the diff amp or you can get big errors. Google should give help on this. Aug 7, 2019 at 9:48