Hundreds of thousands of precision sensing systems have been developed around the world to address the broad spectrum of sensor measurements. These systems, which are used in every major industry for critical monitoring and control applications, are shrinking in size and must consume less power and have higher reliability. And while every application has its own specific set of requirements, many share similar characteristics when it comes to signal conditioning and processing the sensor output.
Independent of sensor type, signals need to go through a conversion process usually provided by an operational amplifier, and then digitized with an analog-to-digital converter (ADC). The choice of components is not a trivial task even for experienced designers. The savvy engineer often requires some level of macro modeling before committing to the evaluation process. Depending on their complexity, these macro models can take several weeks to develop and test before they can be made available to the system designer. In high precision measurements, it is critical to have a thorough error budget analysis to account for all error sources. This process also can be time-consuming and often requires many design revisions.
Texas Instruments has developed a new family of integrated circuits (ICs) called sensor analog front ends (sensor AFEs) that team with WEBENCH Sensor Designer software and a bench-top development platform to provide an easy-to-use solution to the designer's signal-path challenges. This article describes how these configurable sensor AFEs can drastically reduce design time from weeks to minutes when developing temperature, pressure and toxic gas sensing analog front ends.
Overcoming the biggest hurdle: Reducing Design Time
Given that much of the cycle time in precision measurements is spent in the sensor selection and AFE evaluation process, sensor system designers can benefit by using an integrated hardware and software approach that provides a way to quickly address trade-offs, analyze errors associated with the circuit and make instant design changes to achieve the optimum performance. TIís sensor AFEs are complete configurable analog front ends with unique features that work together with WEBENCH software to fast-track signal path designs for a variety of sensors. With WEBENCH, you can to select a sensor, design and configure the sensor AFE and download configuration data to it. The software tool provides real-time estimated device performance for the selected sensor and AFE configuration. In Figure 1, you can see that the LMP90100ís 24-bit sigma-delta ADC is preceded by a programmable gain amplifier (PGA) and a chopper stabilized buffer to eliminate low frequency noise.
The single cycle settling makes the device attractive for multi-channel measurements in pressure and temperature sensors, and the background calibration virtually eliminates offset and gain error over time and temperature. Figure 2 shows the AFE with offset voltage over temperature and calibration turned off, while Figure 3 shows the same sensor AFE with the calibration turned on.
The LMP90100ís sensor diagnostics feature eliminates the need for system debug by automatically detecting faulty conditions, such as determining the presence of an open, short circuit or over range condition. The WEBENCH software speeds evaluation of the LMP90100 by providing hundreds of sensor choices. Various sensors, ranging from thermocouples, RTD, pressure, and load cells are preconfigured and loaded into the WEBENCH tool. If your specific sensor is not available in the list, you can manually add it to the database. After selecting one or more sensors, registers are automatically loaded into the tool and the sensor AFE is automatically configured for that sensor or group of sensors. Youíre now ready to analyze your circuit and apply design changes to the suggested configuration to meet your exact power and performance requirements.
If any device configurations are changed, such as gain or sample rate, the estimated performance table will update automatically and show the new device performance. The software tool lets you learn how the part will meet your needs without having to read a lengthy datasheet.
Advantage over traditional discrete blocks
Historically, sensor interface signal conditioning blocks have used industry standard high voltage op amps to maximize signal-to-noise ratio. In industrial applications such as pressure and temperature sensing, system designers typically choose bipolar input op amps due to their low frequency noise, but if the op amp is interfaced with a high resolution data converter, additional filtering may be needed to attain the desired accuracy. The challenge is then to find a suitable amplifier with a low noise floor, low power consumption, low distortion (usually very high open loop gain) and a robust output necessary to drive the ADC to achieve a cost-effective design and avoid the need for additional external circuitry.
The LMP90100 overcomes such trade-offs by combining a chopper stabilized buffer amp which eliminates flicker noise (1/f noise), a precision PGA and a fourth-order sigma-delta 24-bit modulator with a simultaneous 50/60Hz line rejection digital filter.
A universal approach to gas sensors
One of the most challenging tasks in designing a signal conditioning block for detecting the presence of toxic gases lies in the wide variety of these gases. Depending on which gas you are trying to detect, the concentration can be very different and the design of the AFE is likely to differ in order to accommodate the broad range of these sensors.
Manufacturers of multi-gas sensors and monitors often require the use of several designs, with each design addressing a specific sensing requirement, or meeting a particular price-point depending on the application. With the LMP91000 portable gas detector, designers can use a single IC and the same board to monitor a multitude of gases. The AFEís programmability makes it easy to use across a broad spectrum of gas sensing applications in oil, gas and petrochemical factories. It allows you to accurately monitor carbon monoxide, hydrogen sulfide, sulfur dioxide, chlorine, nitric oxide, methane and much more due to its flexibility and programmable range.
The AFEís architecture (Figure 4) allows it to achieve this flexibility. A voltage divider block is included to prevent the potentiostat circuit (transimpedance amplifier) from an unwanted saturation. The AFE offers the flexibility of programmable gain, and includes a programmable load resistor to isolate the sensor output capacitance from the amplifier.
Flexibility and low current consumption give the sensor AFE a clear advantage over discrete solutions, especially in power conscious portable applications such as wireless sensor networks where multiple sensors communicate together via a central unit. With the WEBENCH software (Figure 5), you can quickly evaluate, analyze and choose the appropriate digitization process before prototyping. You can also combine a MSP430F4152 microcontroller with the LMP91000 to provide a very low power and complete sensor signal path solution.
Development hardware speeds prototyping
Development hardware (see figure 6) that works with the WEBENCH software enables you to investigate the performance of your selected sensor and AFE. This development platform allows you to validate your designs right on the bench. The platform consists of a sensor AFE, an interface to the selected sensor or sensors, and connection to WEBENCH through a PC. You can take data over time, and display the result as a histogram (see Figure 7), or monitor a channel continuously in the oscilloscope-mode, as shown in Figure 8. Data can also be saved to a file and downloaded for performance evaluation using your own software.
The combination of sensor AFE, WEBENCH Sensor Designer software, and development hardware provides an alternative to long hardware development cycles using discrete circuits. Sensor signal path design is simplified and overall design time is dramatically reduced, allowing you to get to market faster with a reliable, proven solution.
About the Author
Soufiane Bendaoud is a product marketing engineer for precision amplifiers at Texas Instruments. Prior to joining TI, Soufiane was a marketing manager at National Semiconductor, where he had worked since 2005. Soufiane received his BSEE degree from San Francisco State University, and an MBA from the University of San Francisco. He has authored more than 20 articles and participated in seminars, trainings and conferences worldwide. In his spare time, Soufiane enjoys playing soccer, the guitar and boxing.