Static vs. Dynamic data acquisition

I have been asked many times to define Static and Dynamic data. In short, the Static data is sufficiently represented just by one sample. Dynamic data is more complex, and is represented by a data set. For example, if we have a sensor that measures temperature, we can get a sample at any time and each time we get a value that represents the temperature at that moment. However, if we sample a sound wave at any time, we will get a random value that is not representative of the sound intensity or frequency.

Data acquisition is the process of converting an analog signal to data by sampling the signal at certain time intervals. The signal is usually an output of a sensor and represents a physical parameter being measured by the sensor. Depending on the type of the sensor, the data can be sampled at a faster or slower rate providing more or less detailed information. A signal from a temperature sensor, for example, changes slowly and can be sampled at a relatively slow rate without loss of quality. The temperature signal is called static as it represents its corresponding value at any time.

A signal from an accelerometer represents a movement of the machine housing on which the accelerometer is installed. This signal contains a sum of vibrations at different frequencies that are generated in different parts of the machine. Analyzing magnitudes of these vibrations helps detect changes in the machine due to wear, lack of lubrication, imbalance, misalignment, and other mechanical, electrical, or process abnormalities. Because of its dynamic nature the accelerometer signal can’t be sampled at an arbitrary time to yield a meaningful result. The data has to be acquired many times in a rapid succession to obtain a data set.

To obtain a proper set of data for analysis the vibration signal has to be sampled at a relatively high sampling rate. The sampling rate defines the analysis bandwidth, i.e. the highest frequency component that is available for analysis. The length of the data set defines the frequency resolution, i.e. the lowest difference between two frequencies that can be detected. The data set then represents a function of vibration magnitude over time. Using further data processing the software can present the data in many other forms, including a spectrum (magnitude vs. frequency) and calculate various overall and statistical parameters.