Using PCs for machine condition monitoring: Part 2

Transducers, signal conditioning, and sensors

By Plant Engineering Staff August 17, 2004

By Brian Betts and Armando Valim, National Instruments, Austin, TX

(Editor’s note: This is the second article in a series on PC-based machine monitoring. The next article in the series will appear in the next issue of Plant Engineering HotWire.)

Plant engineers, maintenance supervisors, and instrument technicians are continually looking for ways to improve reliability and increase equipment uptime. Ensuring that transducer signals or sensor measurements are acquired accurately, reliably, and safely is crucial. Acquiring these signals properly requires signal conditioning.

With the speed and accuracy of modern plug-in data acquisition (DAQ) devices and PLCs, it is easy to overlook the need for signal conditioning. DAQ devices provide analog-to-digital (A/D) conversion, thereby dictating the sampling rate and digital resolution of the system. While they specifically and accurately measure voltage signals, their job is to digitally represent the exact signals at their inputs. The problem is that most sources of inaccuracy contaminate measurements before they are ever digitized. This is where the need for signal conditioning begins.

Nearly 25% of the total cost of developing a DAQ system is spent on system setup. Specifically, installation and configuration of sensors and signal conditioning have traditionally been both time consuming and at risk for user error. New trends and technologies are addressing these potential time sinks by making signal conditioning easier to configure and use. Developers are eliminating the cabling between front-end signal conditioners and the DAQ device performing the A/D conversion by packaging the digitizers in the same chassis as the signal conditioners. In another important trend, developers recently improved measurement and automation system data management by adopting smart sensor technology and increasing the integration of sensor data into the asset management system. Making sensor information an integral part of the asset management system increases plant efficiency, productivity, and reliability.

What is signal conditioning?
Signal conditioning operates strictly in the analog domain, prior to the A/D conversion, to amplify, filter, and isolate measurements. These technologies ensure that the digitized signal accurately represents the signal at the measurement source or sensor placement — even in electrically noisy environments and around hazardous voltage levels.

Using a thermocouple requires signal conditioning. Amplification, filtering, and cold-junction compensation must be provided to accurately measure thermocouple signals.ring must also be provided to eliminate environmental noise that comes from power lines and other electric devices. Cold-junction compensation is also necessary to offset error-causing voltages generated by the Seebeck effect at the connection point to the measurement systems.

The net benefit of these combined signal-conditioning functions is dramatically improved accuracy (Fig. 1). The graph compares thermocouple measurements taken at 25 C using a thermocouple signal conditioning module and a screw terminal connector block with a temperature sensor for cold-junction compensation. The thermocouple signal-conditioning module achieved an accuracy of

Fig. 1 . This screen capture illustrates the difference in accuracy of thermocouple measurements with and without signal conditioning.

In addition to amplification and filtering, many sensors also require electrical excitation. Signal conditioners typically provide this power source. Because different sensor types require current, dc voltage, or ac voltage, signal conditioners commonly are designed for a particular sensor type. Further, some sensors require additional circuitry to facilitate accurate measurements. For example, thermocouples require cold-junction compensation and strain gauges require bridge completion. To accommodate all the features necessary for a specific sensor, signal conditioners are often modular. With modular signal conditioning designs, engineers can plug sensor or signal-specific modules into a common housing or chassis. Modularity enables a single signal conditioning system to measure low voltage, high voltage, electrical current, and a wide variety of sensor types (Fig. 2).

Fig. 2. Modular front-end signal conditioning allows systems to measure low voltage, high voltage, electrical current, and a wide variety of sensor types.

Signal conditioning can also ensure the safety of DAQ systems by providing isolation. Isolation protects both the system hardware and the user from hazardous voltages or voltage spikes by providing a physical barrier between the signal inputs and the PC. Isolation is achieved by converting the electrical signal into another signal type across an isolation amplifier. For example, various types of isolation amps convert electrical signals to magnetic or optical signals before converting them back to a safe, measurable voltage. In addition to ensuring the safety of the system, isolation also breaks ground loops. A ground loop, resulting from current flowing between two system grounds at different potentials, is one of the most common sources of noise in DAQ systems.

Integration of signal conditioning, data acquisition, and USB
Although front-end signal conditioning provides greater flexibility than devices with built-in signal conditioning, it generally takes more time to configure. A DAQ system with front-end signal conditioning requires a user to plug the DAQ device into the PC, cable a chassis to the DAQ device, and install measurement-specific modules in a chassis. Once connected, the user must specify and verify the configuration in software. To simplify this tedious configuration process, DAQ systems now feature new plug-and-play DAQ technologies.

To eliminate cabling between the front-end signal conditioning chassis and the DAQ device, DAQ systems incorporate the analog-to-digital converter (ADC) in the same chassis that houses signal conditioning modules. This gives the system the ease of device configuration with integrated signal conditioning while maintaining the benefits of modular front-end signal conditioners. It also eliminates the need to install plug-in boards in a PC and analog cabling from the PC. In addition to simplifying connectivity, the elimination of long cables carrying analog signals removes a major source of error. Because measured signals are conditioned and digitized in the same chassis, the only connection to the PC is a digital interface to program the system and send data back to the host. To serve this function, the universal serial bus (USB) is emerging as the most widely used data bus. USB is by design a plug-and-play technology, which means that devices connected to the PC are automatically detected and appropriate drivers are automatically installed. This eliminates the requirement to manually specify in software, which devices are installed and which signal conditioning modules are connected.

USB is well suited for DAQ applications. Virtually all desktop and laptop PCs on the market today include one or more USB ports. Cables are standardized for any USB device, including DAQ systems, with standard lengths up to 5 m and available extenders capable of transmission lengths over 500 m. However, the benefits of USB for DAQ systems extend beyond speed and connectivity. Because USB is a plug-and-play technology, it allows the PC to automatically detect when a USB DAQ system is connected. Systems with modular signal conditioning can detect the chassis, modules, and configuration. This eliminates the need to configure jumpers, IRQ settings, and software parameters.

Smart sensors and TEDS
A new standard for sensors, IEEE 1451.4, reduces the time and challenges associated with sensor configuration. The standard establishes a universally accepted method of making sensors plug-and-play — similar to the way a USB mouse is plug-and-play with a computer — and defines a mechanism for adding self-describing behavior to sensors with an analog signal interface. Using any graphical programming language, instrument technicians and operators can take advantage of the benefits of smart sensor technology and integrate it into asset management systems through industry-standard protocols such as XML.

The Transducer Electronic Data Sheet (TEDS) IEEE 1451.4 standard defines the connection of the measurement system to the analog sensor. The standard describes the interface for creating smart sensors with built-in EEPROMS that contain information about the sensor such as calibration and range. A TEDS is essentially an electronic data sheet engineers can use within their test applications. The result is that a system supporting IEEE 1451.4 has plug-and-play sensor features, including automatic uploading of all sensor information, automatic setup of measurement system software and hardware, and automatic use of the calibration information stored in the sensor. This smart sensor functionality eliminates paper data sheets and the manual data entry that they require. Instead, the sensor scaling and calibration information is stored digitally on the sensor itself and is automatically queried by the DAQ system.

The IEEE 1451.4 Mixed-Mode Interface for Smart Transducers standard also defines a mechanism for adding self-identification technology to traditional analog-mode sensors. In addition to providing traditional analog outputs, an IEEE 1451.4 smart sensor also offers a digital interface for communicating with an embedded memory device within the sensor. This memory contains the binary TEDS information that identifies and describes the sensor. The TEDS contains information such as manufacturer, sensor model number, serial number, measurement range, sensitivity, and calibration information (Fig. 3). TEDS templates are designed for the sensor type with which they are associated, and they are specified for a wide range of sensor types.


Fig. 3. Smart TEDS sensors include an embedded TEDS and mixed-mode interface.

There are two types of mixed-mode interfaces defined in the standard — Class 1 and Class 2 interfaces. Class 1 interfaces are intended primarily for constant-current powered piezoelectric transducers (for example, accelerometers and microphones) and define a scheme for sequentially switching between analog mode and digital TEDS mode on a single pair of transducer wires. Constant-current powered transducers, generally referred to as integrated electronics for piezoelectric (IEPE) transducers, incorporate internal signal conditioning powered by a constant current that is sourced by the measurement system on the signal wires. Class 1 transducers take advantage of this de facto analog standard by adding the TEDS with a switch that is controlled by the direction of the current source. By reversing the direction of the current, the instrumentation system switches the sensor into digital TEDS mode.

Most sensor types implement a form of the Class 2 interface, which requires additional wires for digital TEDS communication. The analog I/O of the transducer is left unmodified, and the two-wire TEDS interface is added in parallel to the analog interface. Using this approach, engineers can implement TEDS on virtually any type of amplified or unamplified sensor, including thermocouples, RTDs, thermistors, bridge sensors, electrolytic chemical cells, and 4—20 mA current-loop sensors. In fact, with the add-on approach of Class 2, it is very easy to retrofit existing sensors using a variety of packaging options.

Signal conditioners with TEDS capability include electronics to read and write to the TEDS chip on the sensor. When connected, a sensor can be automatically detected and tell the DAQ system information about its type and scaling requirements. This eliminates the need to manually enter information from a sensor’s paper data sheet into software. Smart TEDS sensors prevent the need to keep track of which sensor is connected to which channel. Regardless of which channel it is connected to, the TEDS sensor can tell the system how to properly measure and scale it.

Conclusion
Together, USB and TEDS technologies are reshaping the way engineers use signal conditioners for sensor measurements. From the sensors to the signal conditioning modules and digitizer, the entire system offers engineers the benefits of plug-and-play technology. Manual system configuration, data entry, and lost paper data sheets are things of the past. Signal conditioning is still required to ensure safe and accurate measurements, but its configuration and use is now easier than ever before.

Brian Betts works as a DAQ Systems product marketing group manager at National Instruments, Austin, TX He holds a B.S. in mechanical engineering from Texas A&M University. For more information about signal conditioning, TEDS, and smart sensor technology, Brian can be reached at 512-683-8711 or brian.betts@ni.com .

Armando Valim is a group manager responsible for sound and vibration products at National Instruments, Austin, TX. He holds a bachelors in engineering as well as graduate degrees in engineering, marketing, and finance. For more information about using PC-based machine monitoring, Armando can be reached at 512-683-5861 or armado.valim@ni.com .


Related Resources