Signal Analysis and Intelligent Flowmeters

Abstract

In general, the instruments used for process measurement provide an acceptable level of performance, but there are occasions when the process operators suspect the integrity of the measurement signal and feel obliged to call on the instrument staff to check the span and zero of the transmitter. Quite apart from the cost of making such a test, experience shows that many of the transmitters checked in this way, and shown to be functioning correctly, subsequently become defective. Even if it were possible to devise a remotely controlled means for checking the span and zero, it is doubtful whether it could ever become a practical proposition. An entirely different approach is to use active sensors, such as resonant elements, in place of the traditional passive devices and to apply signal analysis techniques, which can be implemented by microprocessors. The measurement signal for these and other types of sensor can be derived from the time interval between successive zero crossings and this avoids the need for A/D conversion. Furthermore, analysis of the signal enables information regarding the status of the sensor as well as the operation of the associated process plant to be determined. This has been demonstrated by the application of these SFT to the signals generated by a turbine flowmeter.