When are signal conditioners not needed with sensors?

Signal conditioning is necessary in most sensor applications, but there are exceptions. This article discusses those exceptions and reviews sensor technology, system design, and operational considerations where signal conditioning may be unnecessary and expensive.

Sensing is all about the output signal. External signal conditioners are not needed when signal conditioning is integrated into the measurement device. Signal conditioning is often unnecessary for sensors with large output signals or for making simple measurements in environments with minimal noise. It may not be needed with passive sensors like photodiodes, thermocouples, tunnel magnetoresistance (TMR) sensors, and on/off sensors like level floats, reed switches, and proximity/limit switches.

To understand why signal conditioning may not be needed, it’s useful to start with a review of the basic structure of a sensor circuit:

Sensing elements typically produce a low-level signal subject to interference from various noise sources.
Signal conditioning circuits filter the noise, amplify the clean signal, and may provide further processing like analog-to-digital conversion.
Communication interfaces are sometimes included as part of the signal conditioning process since they prepare the sensor for connection with the overall system.

Some sensors are delivered in a single package that includes the sensing element, signal conditioning, and communication interface, eliminating the need for external signal conditioning (Figure 1). However, not all sensors or sensor applications require signal conditioning.

signal conditioners — Figure 1. An example of a sensor design that produces a digital output does not need additional signal conditioning. (Image: Monolithic Power Systems)

Factors related to the need for signal conditioning

Signal conditioning is generally unnecessary when the sensor output is already within the voltage range and signal quality (low noise) required by the data acquisition system (DAQ) and when its impedance matches the DAQ.

Other factors to consider include:

If the sensor produces a strong, clean signal, conditioning may not be needed
If the measurement is simple and low-precision in a low-noise environment
Some sensors naturally produce a digital output and don’t need signal conditioning.
Basic detection applications like the presence or absence of an object, exceeding a specific threshold, or monitoring for simple changes in state may not benefit from signal conditioning.

TMR sensors

The output of a TMR sensor can be 500 times higher than a Hall effect sensor, 20 times higher than that of an anisotropic magneto-resistive (AMR) sensor, and 6 times higher than a giant magnetoresistance (GMR) sensor. In addition, TMR sensors can have high signal-to-noise ratios, making signal conditioning unnecessary, depending on the specifications of the DAQ or microcontroller that reads the sensor signal.

TMR output can be compatible with an analog-to-digital converter’s (ADC) input impedance and voltage range , making signal conditioning unnecessary in applications where high precision is not critical.

The resistance changes continuously when a magnet is rotated on a TMR sensor. Since the resistance value is proportional to the relative angle between the magnetization directions of the pin layer and the free layer, it can be utilized as an angle sensor (Figure 2).

It’s necessary to carefully consider operating conditions when deciding if signal conditioning is needed with a TMR sensor. Considerations include:

Low noise environments may reduce or eliminate the need for filtering. But filtering may be needed even in low-noise environments to support high-precision applications.
Applications subject to large temperature swings can benefit from temperature compensation circuitry.
Linearization of the sensor output may be needed to increase precision.

Float level sensors

Basic float-level sensor applications don’t need signal conditioning. The sensor provides a strong and simple on/off output that the host system can easily read.

Some filtering may be needed to eliminate false readings in especially noisy environments. Float-level sensors might require amplification to match the input range of a DAQ. Signal conditioning may also improve sensor resolution in high-precision applications.

Summary

Signal conditioning is usually necessary for sensor applications, but not always. Factors like simple measurement applications, benign and low-noise environments, and sensors with large output signals that are well-matched to the DAC can eliminate the need for signal conditioning. Careful analysis of all relevant factors is necessary to determine whether signal conditioning is needed or an unnecessary complication and expense.