Sensor Signal Conditioning, Bridge Measurement, Calibration, and ADC Scaling

Sensor design is a translation problem. A physical quantity such as force, pressure, current, optical power, or acceleration must become an electrical signal, then a calibrated engineering value. Most mistakes happen when a designer skips one layer of that translation: treating sensitivity as amplifier gain, ignoring offset, mixing units, or assuming ADC resolution equals accuracy.

This guide explains the shared engineering model behind ECParts sensor calculators. It focuses on sensor signal conditioning, bridge measurement, calibration slope and offset, ADC scaling, and practical measurement limits. Motion-specific geometry and time-of-flight detail are intentionally kept as supporting context because those deserve a later motion and distance sensor guide.

Sensitivity is the slope between input and output. It is not a unitless gain unless both sides use compatible electrical units. A load cell may specify mV/V, a Hall sensor mV/A, an accelerometer V/g, and a photodiode A/W. Always carry units through the calculation.

Offset is the output at a reference input. Span is the difference between maximum and minimum values. Range names the endpoints. Gain correction changes slope; offset correction shifts the zero point. A robust sensor workflow keeps these separate.

Strain gauges and load cells often use Wheatstone bridges because tiny resistance changes are easier to measure as differential voltage ratios. A bridge output in mV/V is not an absolute voltage by itself. It scales with excitation voltage, which makes excitation accuracy and ratiometric measurement important.

Bridge sensor formulas are owned by the Sensors category because the engineering task is measurement. They are related to resistor dividers, but the important result is sensor output polarity, sensitivity, zero balance, bridge excitation, and measurement range rather than generic resistor-network design.

Many sensor calculators use a linear model because it is practical and transparent. Two endpoints define slope and intercept, and a measured signal is converted into engineering units. This model is useful for pressure transmitters, Hall current sensors, accelerometer voltage outputs, generic analog sensors, and calibrated raw readings.

Calibration improves a measurement model under the conditions used during calibration. A one-point tare can reduce offset. A two-point calibration can correct slope and offset for a linear model. A multi-point fit can estimate residuals. None of these automatically remove noise, hysteresis, drift, nonlinearity, or installation error.

ADC resolution connects the conditioned sensor span to digital codes. The useful question is not only how many bits the ADC has, but how much of the input span the sensor actually uses after gain and offset. A 12-bit converter gives 4096 possible codes, but a sensor that uses only half the ADC range gets about half the code span for measurement.

Resolution is not accuracy. ADC reference tolerance, sensor nonlinearity, noise, amplifier offset, input leakage, source impedance, anti-alias filtering, and temperature drift can all be larger than one ideal code width.

Sensors V1 does not take ownership of NTC thermistors, RTDs, thermocouples, or temperature coefficient models. Those remain Thermal topics because their primary equations are temperature conversion, resistance-temperature behavior, and thermal sensing limits.

Op-amps are part of many sensor front ends, especially bridge amplifiers and photodiode TIAs, but op-amp gain, stability, slew rate, and feedback topology remain Op-Amp topics. Use Op-Amp Feedback, Gain, and Signal Conditioning Basics when the active circuit becomes the main question.

Digital and converter topics also have boundaries. ADC code width is discussed here only as sensor measurement resolution. Pure base conversion belongs to converter tools, while logic states and synchronous timing belong to the Digital Logic Fundamentals Guide.