01 May 2003
Plug-and-play sensors make big promises
New standard benefits go deeper.
By Brent Boecking
The IEEE smart sensor standards, first introduced in the mid-1990s, contained grandiose frameworks of sophisticated networking technologies and onboard DSP-enhanced intelligence. But an emerging IEEE smart sensor standard, IEEE P1451.4, takes a more pragmatic approach to smart sensing than its predecessors and promises to deliver a broad and significant impact.
The standard delivers on the most crucial feature of smart sensing technology: self-identification of the sensor. Compared with the past smart sensor standards, IEEE P1451.4 plug-and-play sensors are inexpensive, simple to develop, and compatible with existing measurement systems.
Plug-and-play sensors consist of two major components: the traditional analog sensor and electronics that contain information about the sensor. The simple yet powerful difference between traditional analog sensors and plug-and-play sensors is transducer electronic data sheets (TEDS)—onboard electronics containing model-specific information such as manufacturer name, model number, and serial number, as well as specific information such as calibration data. TEDS electronics are inexpensive memory components, typically electrically erasable programmable read-only memory (EEPROM), which communicates digitally to the data acquisition system.
The digital communication can occur over the same wires as the analog signal or over separate, dedicated wires. TEDS contains all of the critical information necessary to describe how to configure the data acquisition system, both in hardware and software, to present calibrated data in the correct engineering units. You can apply TEDS to a wide variety of sensor types: load cells, integrated electronics piezoelectric accelerometers, thermocouples, linear variable differential transformers, resistance temperature detectors, and any other general analog output sensor.
Two sections within each TEDS must be present for plug-and-play capability: basic and standard. The information fields within basic TEDS are the same for all sensors: model and manufacturer ID. Standard TEDS information fields, however, depend on the sensor's classification. The proposed standard says 44 sensor types have their own standard TEDS template, which contains technical information such as measurement range, minimum and maximum excitation, and electrical output, permitting the data acquisition system to properly configure itself and present the data.
You can write two more sections to the TEDS—calibration and a user area—to accommodate the specialized needs of some sensors. The calibration TEDS can contain polynomial curve-fit information or lookup tables describing sensor output over the entire operating range. The user area can store sensor location, sensor owner, or any type of custom information.
Data acquisition (DAQ) hardware that can read IEEE 1451.4 plug-and-play sensors can also take data from existing legacy sensors. In addition, the TEDS can be accessed from the DAQ hardware or can be accessed as Virtual TEDS via the internet, making any sensor plug-and-play.
TEDS DIGITAL INTERFACE
IEEE P1451.4 describes TEDS as physically associated with the sensor, either as part of the sensor itself or as part of the cable assembly. To minimize the size of the TEDS, a low-mass memory device stores data in binary format—typically an EEPROM. The memory devices use a simple, low-cost serial transmission protocol called 1-Wire. The 1-Wire protocol uses two conductors. The first conductor provides power from the data acquisition system to the sensor as well as a transmission path for the serial data. The second conductor provides a ground reference. These EEPROMS come in a variety of footprints and memory capacities.
You can interface to an IEEE P1451.4 sensor by two methods. The first method, a Class 1 interface, uses the same conductors for the digital and analog signal. Use this interface with constant-current powered sensors such as piezoelectric accelerometers and microphones. The second method, a Class 2 interface, uses separate conductors for the digital and analog signals. Most sensors use the Class 2 interface to transfer data to the data acquisition system, leaving the analog portion of the transducer unmodified. That's why nearly any sensor can benefit from IEEE P1451.4.
Today's Internet technology will allow the TEDS data structure to bring immediate benefits to traditional analog sensors through virtual TEDS. Virtual TEDS contain the same binary information located in physical TEDS; however, the information is located in Internet-ready databases available to anyone configuring a sensor. With this virtual instrumentation approach to sensing, engineers and scientists will realize the benefits of today's plug-and-play sensors. At the same time, sensor and instrumentation vendors will race to deliver IEEE P1451.4–compliant products to the marketplace.
Ease of use and automatic configuration are just some of the benefits of plug-and-play sensors, especially in the case where the cost of configuring thousands of sensors far surpasses the cost of the sensor itself. Measurement accuracy and manufacturing cost will advance with this new approach to sensing technology.
Today's analog sensor designs function with antiquated, unintelligent instrumentation, forcing trade-offs during design and manufacturing. Most sensors today should be completely linear within their operating range or must adhere to strict standard polynomial curve fits. These constraints allowed the data acquisition system to convert the raw analog inputs to physical units and present them to the user. Thermocouples typically provide a limited temperature range and require time-consuming manufacturing techniques to create metal junctions with properties that adhere to the thermocouple standards (such as K, J, and S) established in the early 1900s.
Computer-based data acquisition and presentation provide a solution to traditional sensor technologies' limitations. We don't have to adhere to standard polynomial responses if we provide sophisticated analysis and postprocessing of data in software. Instead, we can optimize sensors for other factors such as cost, stability, or manufacturability, and we can easily characterize the behavior of the device and compensate for it in software.
MANUFACTURER IMPROVES SENSORS
Watlow Electric Manufacturing Co., which manufactures temperature measurement sensors, has made an accurate thermocouple out of a clothes hanger and bailing wire to demonstrate the effectiveness of plug-and-play technology applied to sensors. Ordinarily, with antiquated, unintelligent instrumentation, this thermocouple would be useless because it does not adhere to a standard letter designation. Plus, its natural voltage output is not linearly proportional to the environmental temperature. However, by adding a TEDS, which contains a custom lookup table that correlates voltage output to temperature, manufacturers have made this odd metal combination serve as an accurate thermocouple.
When Watlow applies IEEE P1451.4 technology to a broad set of its products, the company significantly lowers its manufacturing cost. The company doesn't design its thermocouples according to aging standards, which can result in scrapping complete lots of raw material. By using new high-temperature alloys, the company can provide better high-temperature drift performance.
Other sensors—load cells, LVDTs, and RTDs—can apply the same approach, Watlow said. IEEE 1451.4 sensors' self-describing behavior not only improves end users' ability to configure and use the sensors but also lowers cost and boosts performance.
Because IEEE P1451.4 (the proposed standard) is based on off-the-shelf technology, the commercial industry has already adopted it. However, developers continue to work on the IEEE P1451.4 proposed standard and expect to ballot and ratify it early this year. IT
Behind the byline
Brent Boecking is a marketing engineer for data acquisition at National Instruments in Austin, Texas.
Return to Previous Page