This design guide describes the signal chain of common sensor transmitters for pressure, temperature, current, optical signal detection, and proximity detection. The article introduces the intricacies of signal channel selection. The design examples and block diagrams provided will help readers choose the best components to meet the different needs of the system.

<-- ============================================================================ --><-- CONTENT: DB HTML --><-- =========================================================================================================== -->

Pressure Sensors and Weigh Scales (Load/Sensing)

Overview

In modern industrial control and system monitoring, pressure and weight are often monitored and measured. Pressure measurement is particularly important because pressure can be used directly to measure fluids, heights, and other physical quantities. Since load is a property that affects the output of the sensor, pressure and weight measuring devices can be considered "load sensors." Load sensors are used in a wide range of applications, from vacuum gauges to heavy machinery weighing, industrial hydraulic equipment, and absolute pressure sensors (MAP). Each application has different specific requirements for precision, accuracy, and cost.



Although there are many methods and technologies for measuring pressure and weight (load/sensing), the most commonly used measurement device is a strain gauge.

The two most common strain gauges are metal foil, which is mostly used in weight/pressure sensors, and semiconductor-based piezoresistive sensors, which are widely used for pressure measurement. Compared with metal foil sensors, piezoresistive sensors have higher sensitivity and better linearity, but are easily affected by temperature and have a certain initial deviation.

In principle, all strain gauges change resistance when subjected to external forces. Therefore, when stimulated by an electrical signal, they can effectively convert pressure and weight into electrical signals. Typically, one, two, or four of these active resistor elements (strain gages) are placed on a Wheatstone bridge (sometimes called a load cell), which produces a differential output voltage that corresponds to pressure or weight.

Engineers can design a sensor module that can meet the needs of a variety of load/sense systems. A successful design requires a sensor element to detect the physical quantity and a properly designed signal chain.


Block diagram of the signal chain for a load/sense system. For more information on Maxim's recommended pressure sensor solutions, visit www.maxim-ic.com/psi.

Complete signal chain solution

The sensor signal chain must be able to handle weak signals with noise. In order to accurately measure the change in the output voltage of a resistive sensor, the circuit must have the following functions: excitation, amplification, filtering, and acquisition. Some solutions may also require the use of digital signal processing (DSP) technology to process the signal, error compensation, digital amplification, and user-programmable operations.

excitation

A high-precision, stable voltage or current source with very low temperature drift is often used as sensor excitation. The sensor output is ratiometric to the excitation source (often expressed in mV/V). Therefore, the analog-to-digital converter (ADC) and excitation circuit are usually designed to use a common reference, or the excitation voltage is used as the reference of the ADC. Additional ADC channels can be used to accurately measure the excitation voltage.

Sensor/Bridge

This portion of the signal chain functions as the strain sensor, which is placed in the load cell (Wheatstone bridge design) section as described above in the “Overview” section.

Amplification and level shifting—Analog front end (AFE)

In some designs, the sensor output voltage range is very small and requires resolution in the nanovolt range. In this case, the sensor output signal must be amplified before it is sent to the ADC input. To prevent errors introduced during the amplification stage, a low noise amplifier (LNA) with low offset voltage (V _OS ) and low temperature drift needs to be selected. The disadvantage of the Wheatstone bridge is that the common-mode voltage is much larger than the useful signal. This means that the LNA must also have a very high common-mode rejection ratio (CMRR), typically greater than 100dB. If a single-ended ADC is used, additional circuitry is required to eliminate the high common-mode voltage before data acquisition. In addition, due to the narrow signal bandwidth, the 1/f noise of the amplifier will also introduce errors. Therefore, it is best to use a chopper-stabilized amplifier. Using a very high resolution ADC and occupying a small portion of the full-scale range can help to relax the stringent requirements on the amplifier.

Acquisition—ADC

When selecting an ADC, it is important to carefully check its specifications, such as noise-free range or effective resolution, which indicates the ADC's ability to discern fixed input levels. An alternative is noise-free counts or codes. Most high-precision ADC data sheets present these specifications as a table of peak noise or RMS noise versus speed, or sometimes as a noise histogram.

Other ADC specifications to consider include low offset error, low temperature drift, and excellent linearity. For certain low-power applications, the relationship between speed and power consumption is also a very important specification.

Filtering

Sensor signals are typically narrow bandwidth and highly sensitive to noise. Therefore, limiting the bandwidth of the signal through filtering can significantly reduce the overall noise. Using a Σ-Δ ADC can simplify the noise filtering requirements because this architecture provides inherent oversampling characteristics.

Digital Signal Processing (DSP)—Digital Domain

In addition to analog signal conditioning, the acquired signal needs to be further processed in the digital domain to extract the signal and reduce noise. It is usually necessary to find algorithms that are specific to the specific application and its nuances. Some common algorithms, such as offset and gain calibration in the digital domain, linearization, digital filtering, and compensation based on temperature (or other constraints).

Signal Conditioning/Integrated Solutions

Some integrated solutions integrate all the required functional modules into a single chip, usually called a sensor signal conditioner IC. A signal conditioner is an application-specific IC (ASIC) that compensates, amplifies, and calibrates the input signal and can cover a wide temperature range. Depending on the different accuracy requirements of the signal conditioner, the ASIC will integrate all or part of the following modules: sensor excitation circuit, digital/analog converter (DAC), programmable gain amplifier (PGA), analog/digital converter (ADC), memory, multiplexer (MUX), CPU, temperature sensor, and digital interface.

There are two common types of signal conditioners: conditioners for analog signal paths (analog conditioners) and conditioners for digital signal paths (digital conditioners). Analog conditioners have a faster response time and provide a continuous output signal that reflects the real-time changes of the input signal. They usually use hardware compensation mechanisms (not flexible enough). Digital conditioners are often based on microcontrollers and have a slower response time due to the execution time of ADC and DSP algorithms. The resolution of the ADC should be considered to minimize the quantization error. The biggest benefit of digital signal conditioners is that they provide flexible compensation algorithms that can be adjusted according to the user's application.

Temperature detection

Overview

The main role of temperature detection in industrial systems is manifested in three aspects.

1. Temperature control , such as constant temperature ovens, freezers and environmental control systems, implements heating/cooling operations based on measured temperature judgment.
   
   
2. Calibration Various sensors, oscillators, and other components often vary with temperature. Therefore, it is necessary to measure temperature to ensure the accuracy of sensitive system components.
   
   
3. Protect components and systems from damage at extreme temperatures. Temperature detection determines the appropriate measures to be taken.

Thermistors, RTDs, thermocouples, and ICs are the most widely used temperature sensing technologies today. Each design has its own advantages (e.g., cost, accuracy, temperature range) and is suitable for different specific applications. Each of these technologies will be discussed below.

In addition to providing the industry's most comprehensive range of dedicated temperature sensor ICs, Maxim also offers any device needed to interface a system with thermistors, RTDs, and thermocouples.