IMPACT OF NETWORKED DATA ACQUISITION SYSTEMS ON TRANSDUCERS

Abstract

Transducers have traditionally been incorporated into data systems by connecting the transducer to a signal conditioner that is then connected to a multiplexer with an Analog-to-Digital Converter (ADC). The signal conditioning, multiplexer and the ADC are usually included within the same assembly that is called a Data Acquisition Unit (DAU) or an encoder. A network centric data system allows the same architecture to be used if the interface to the encoder is changed to be a network interface. However, a network centric architecture allows other options as well. The signal conditioning and ADC can be included within the same package as the transducer and the assembly can be interfaced to the network. When this is combined with the processing capability now available, a whole new range of possibilities present themselves. The transducer can now be digitally processed to provide a linear output, it can be converted to Engineering Units, digitally filtered or have a host of other functions performed within the housing that contains the transducer. However, the network centric approach does not produce these advantages without some disadvantages. The major problem that needs to be solved is how we time stamp the data. With the encoder we could time stamp the PCM frame and be able to determine the time that a sample was taken from that information. Even in systems that convert the encoder to have a network interface, the time stamp needs to be affixed to the data in the encoder. With a network centric approach, the sample can be taken in the transducer and how to time stamp it becomes a real problem. This is a problem that must be considered at the system level. Some method of making time available at a low enough level in the system to allow transducer outputs to be time stamped is either a network issue or it requires a separate interface.