When people hear the words “data acquisition”, they often directly think of big data used by data analysts in the business world. However, in the field of science and engineering, data acquisition systems refer to collecting data from the real world to determine its physical conditions.
That said, data such as temperature, current, water level, weather, wind direction are some of what data acquisition systems (DAQ) deal with. To start the intro to DAQ Systems, let us briefly look into how DAQ systems.
Data acquisition systems convert signals into digital values which a computer can read and interpret. The process starts with placing the data acquisition device in its desired location; this electronic device contains sensors that turn physical parameters into electrical signals. These sensor signals are then processed by a signal-conditioning circuitry which convert the current signals into digital values. Lastly, these conditioned signals are then sent to analog-to-digital converters which does the final conversion into digital values appropriate for the computers to read.
This refers to the number of measurement values that a data acquisition system can provide.
Most commonly, if not universally, resolution is listed as bits. With the help of input range, the resolution can tell the smallest change detectable in the input.
For applications that require low sample rates but high resolution, 20-bit and 24-bit are commonly used; whereas, for applications requiring high-speed resolution, 8-bit converters are usually preferred.
A common misconception is that accuracy and resolution are just one specification; they are actually different from one another. Even though a DAQ system can resolve a signal, it does not necessarily mean that the inputs are accurate. Accuracy is actually one of the most important specification to every data acquisition system.
The number of channels is one of the clearest specification to consider when choosing a data acquisition system. The number of channels will depend according to the number of signals you wish to measure.
Other than the number of channels, it’s also important to determine if the system has both differential and single-ended inputs. In addition, try to determine if there are any hidden analog inputs in that system.
An error that will always be present in every data acquisition system is noise. Usually, the noise is generated from the external system and often from cabling. For every data acquisition system, as well, there is an internal noise. The noise is mostly measured by acquiring a series of samples by shorting the inputs at the device connector.
- Non linearity
Non linearity refers to the difference between the input or plotted measurement, the ideal measurement and the actual voltage. This error has two components; the differential nonlinearity and the integral non linearity.
- Input offset
The input offset refers to the constant difference between the measured input and the actual input voltage, assuming all other errors are zero.
- Grain error
Assuming there are no other errors, gain error refers to the difference in the slope in volts between the ideal system and the actual system. However, in the real world where there are other errors, gain error refers to error in the measurement during the full-scale reading.