What is signal conditioning? Why is it used in data acquisition and data logging?
A strain gauge on its own does not transmit any form of signal, it that merely has a resistance in its circuit, which varies as the gauge is distorted when the object to which it is bonded is placed under stress or strain.
What signal conditioning does essentially is puts a current through the sensor so that it delivers a very low level signal, and then amplifies and calibrates that into readable units.
If the gauge is part of the simplest type of stress analysis circuit it forms a quarter-bridge, one leg of what’s called a Wheatstone bridge. So a quarter-bridge is one strain gauge measuring strain, either axial or bending, as one leg of a Wheatstone bridge circuit which returns a low level signal in millivolts when energised.
The instrumentation creates this Wheatstone bridge circuit by taking the quarter bridge, adding in resistors of a known resistance, putting voltage through the circuit, taking in the returned low level signal and amplifying it by known values and including a calibration facility, so that data is outputted in engineering units, not just millivolts.
In the case of both data acquisition and data logging this provides what engineers need. They just want to know what stress is being experienced at point A and what strain is being experienced at point B measured in the units they work in.
All a strain gauge provides is a resistance, or if you energise it as part of a Wheatstone bridge without signal conditioning, it returns millivolts. Most engineers can’t interpret these forms of data, or rather haven’t got time to. So the signal conditioning instrumentation does all the work of making the signal readable, and recordable.
Signal conditioning thus completes the Wheatstone bridge, energises the bridge, then amplifies the signal and applies shunt calibration to it, outputting information in known physical units, and recording that information.
Signal conditioning works in the same manner with more complex half-bridge and full-bridge circuits, where the Wheatstone bridge contains two or four strain gauges for greater accuracy and sensitivity.
There are many data loggers available, but these only record an output in millivolt units.
They won’t put current through the bridge, won’t complete the quarter bridge, they just take in millivolts. So if you can’t energise the bridge, you won’t get millivolts out, so a data logger can’t record it.
This is similar to using a linear variable displacement transducer (LVDT). It needs energisation with an AC carrier wave, and needs special signal conditioning to convert that to a voltage output. Similarly an eddy current probe or a capacitance proximity detector, these all require some sort of conditioning interface to use them.
Thus, when acquiring data, or logging it, most of the available sensors need a special interface to be able to record data that can be interpreted easily.