Electronics, Signal Conditioning and Data Acquisition

Introduction

Mechatronic systems connect physical processes to computation and action. Sensors convert mechanical, thermal, optical, magnetic, or chemical variables into electrical signals; electronic interfaces condition those signals; data-acquisition hardware converts them into numerical data; and controllers use the data to command actuators. The quality of this chain determines whether a system is accurate, stable, safe, and reliable.

A sensor rarely produces a signal that can be connected directly to a microcontroller. Its output may be only a few millivolts, contain unwanted noise, have a high source impedance, or vary outside the input range of an analog-to-digital converter (ADC). Signal conditioning solves these problems through amplification, filtering, linearization, isolation, level shifting, and protection. Digital interfaces and switching circuits then allow the controller to communicate with sensors and drive loads.

This chapter develops the principles required to design basic analog and digital interfaces for mechatronic applications. It emphasizes practical design reasoning: defining signal requirements, selecting suitable circuits, checking loading and bandwidth, protecting electronics, and controlling noise.

Learning Objectives

By the end of this chapter, a student should be able to:

1. Explain the role of signal conditioning and data acquisition in a mechatronic system.
2. Analyze voltage dividers, bridges, operational-amplifier circuits, filters, and switching interfaces.
3. Select ADC and DAC parameters using resolution, range, sampling, and quantization considerations.
4. Design basic protection, grounding, shielding, and noise-mitigation arrangements.
5. Evaluate an interface against accuracy, bandwidth, safety, loading, and environmental requirements.
6. Apply these principles to a practical sensor-to-controller system.

1. The Measurement and Control Signal Chain

A typical chain is:

Physical variable → sensor → signal conditioning → ADC or digital interface → processor → DAC, PWM, or switching stage → actuator → physical process

The signal-conditioning block may include:

• excitation for resistive or bridge sensors;
• amplification of low-level signals;
• attenuation or level shifting;
• low-pass, high-pass, band-pass, or notch filtering;
• rectification or demodulation;
• electrical isolation;
• overvoltage, overcurrent, and electrostatic-discharge protection;
• conversion between voltage, current, frequency, and digital forms.

A successful design begins with a signal specification. Important quantities include minimum and maximum signal, desired accuracy, bandwidth, common-mode voltage, source impedance, supply voltage, environmental noise, cable length, and fault conditions. A circuit that is electrically functional may still be unsuitable if it saturates, aliases high-frequency noise, loads the sensor, or lacks protection.

2. Basic Analog Interface Concepts

2.1 Voltage dividers and loading

Two resistors form a voltage divider. With input voltage Vᵢ applied to series resistors R₁ and R₂, and output measured across R₂,

Here Vᵢ and Vₒ are in volts and R₁ and R₂ are in ohms. The equation assumes ideal resistors and that the output is unloaded. If a load resistance R_L is connected, the effective lower resistance becomes R₂ ∥ R_L:

and R₂ in the divider equation must be replaced by R₂,eff. This loading effect is important when a sensor or divider is connected to an ADC with finite input resistance.

A voltage divider is often used with a thermistor or potentiometer. Its limitations include dependence on resistor tolerance, supply variation, self-heating, and nonlinear sensor behavior.

2.2 Operational amplifiers

An ideal operational amplifier (op-amp) is modeled with infinite open-loop gain, infinite input resistance, zero output resistance, and unlimited bandwidth. Real devices have finite gain-bandwidth product, input offset voltage, bias currents, noise, output-current limits, and input/output voltage restrictions.

With negative feedback and operation within the linear region, two useful approximations are:

• input currents are approximately zero;
• the two input voltages are approximately equal.

These approximations do not apply during saturation, without negative feedback, or when frequency-dependent effects become significant.

For a non-inverting amplifier,

where R_f is the feedback resistor and R_g is the resistor from the inverting input to the reference node. The voltage gain is dimensionless. For an inverting amplifier,

where Rᵢ is the input resistor. The negative sign indicates phase inversion relative to the reference node.

A differential amplifier rejects signals common to both inputs. Its common-mode rejection ratio is

where A_d is differential gain and A_cm is common-mode gain. CMRR is often expressed in decibels:

A high CMRR is valuable when a small sensor signal is superimposed on a large common-mode voltage. Instrumentation amplifiers are precision differential amplifiers with high input impedance, accurately set gain, and strong common-mode rejection. They are commonly used with strain gauges, load cells, and biomedical or industrial sensors.

2.3 Offset, saturation, and headroom

If a single-supply ADC accepts 0 to 3.3 V but a sensor produces −20 to +20 mV, amplification alone is insufficient because the negative half-cycle cannot be represented. A reference or bias voltage can shift the signal:

where V_ref is a bias voltage and G is the signed gain. The design must satisfy the complete output range, including offset, tolerance, noise, and transient margin. The op-amp must also support the required input common-mode range and output swing at the chosen supply voltage.

3. Bridges and Small-Signal Sensors

A Wheatstone bridge converts small resistance changes into a differential voltage. For four bridge resistances R₁, R₂, R₃, and R₄ excited by V_ex, with output measured between the two divider midpoints,

The sign depends on which midpoint is defined as positive. At balance, the two ratios are equal and Vₒ = 0. For a quarter-bridge with three equal resistances R and one resistance R + ∆R, the exact expression depends on resistor placement. If the changed resistor is in one arm and |∆R| ≪ R, the magnitude of output is approximately

This small output requires a differential or instrumentation amplifier. Bridge accuracy is affected by resistor matching, excitation stability, temperature gradients, lead resistance, amplifier offset, and mechanical hysteresis. Ratiometric ADC measurement—using the same excitation reference for the bridge and ADC—can reduce sensitivity to excitation-voltage variation.

4. Filters and Bandwidth

Filters pass desired frequency components and attenuate unwanted components. A first-order RC low-pass filter consists of a series resistor R and shunt capacitor C. Its transfer-function magnitude is

where ω = 2πf is angular frequency in radians per second. Its −3 dB cutoff frequency is

The formula assumes ideal components, a sufficiently high load impedance, and negligible parasitic effects. At f_c, the output magnitude is 1/√2 ≈ 0.707 of the low-frequency value. Above the cutoff, a first-order low-pass filter attenuates at approximately 20 dB per decade.

A high-pass filter can be formed by exchanging the series and shunt positions of the capacitor and resistor. It blocks DC and passes sufficiently high frequencies. Multiple stages provide stronger attenuation but may interact unless their loading is controlled. Active filters use op-amps to provide gain, buffering, and more precise frequency responses.

Filtering must be designed from the signal bandwidth, not merely from the noise frequency. Excessive filtering introduces amplitude attenuation and phase delay. In feedback control, phase delay can reduce stability margins. In measurement, a filter that is too slow may conceal genuine transients.

Anti-aliasing

Sampling converts a continuous-time signal into discrete data. If frequency components above half the sampling frequency are present, they can appear as false lower-frequency components. The Nyquist frequency is

where f_s is sampling frequency. A practical system uses an analog anti-aliasing low-pass filter before the ADC and samples sufficiently faster than the highest signal frequency of interest. The Nyquist condition is necessary but not always sufficient because a real filter has a transition band rather than an ideal cutoff.

5. Analog-to-Digital Conversion

An ADC maps an analog input into a finite set of digital codes. For an ideal unipolar N-bit ADC with input range 0 to V_ref, the number of codes is 2ᴺ and the nominal code width, often called one least significant bit (LSB), is

Some data sheets define the ideal full-scale endpoint differently, so the exact code-transfer convention must be checked. The value q is a voltage and represents the ideal quantization interval. Quantization error is commonly modeled as approximately ±q/2 for rounding, assuming an input that is not correlated with the quantization process.

For an ideal ADC, an input V_in is converted approximately to

for a unipolar converter using endpoint coding, where floor means retain the integer part. Actual converters have offset error, gain error, integral nonlinearity, differential nonlinearity, noise, finite acquisition time, and reference uncertainty.

Worked example: ADC resolution

A 12-bit ADC uses a 0–3.3 V reference. The ideal number of codes is 2¹² = 4096, so

An input of 1.65 V is approximately half-scale and produces a code near 2048, subject to the converter’s coding convention and nonideal errors. If a sensor signal spans only 0–0.33 V, using the full ADC range directly wastes most available codes. A gain of approximately 10, provided the sensor and amplifier permit it, uses more of the converter range.

ADC sampling and acquisition

The source driving an ADC’s sample-and-hold capacitor must settle within the required acquisition time. A large source resistance may cause conversion error even when the ADC has high nominal input impedance. A buffer amplifier, a suitable RC network, or a longer acquisition time may be required. The capacitor and source resistance form a time constant:

where R_s is effective source resistance and C_s is the sampling capacitance. The design should be checked against the ADC data sheet rather than relying solely on the simplified time constant.

6. Digital-to-Analog Conversion and PWM

A DAC produces an analog output from a digital code. For an ideal unipolar N-bit DAC with reference V_ref, the approximate output for a normalized code D is

where D ranges from 0 to 2ᴺ − 1. Some DAC architectures use a different full-scale convention; the manufacturer’s transfer equation takes precedence.

Pulse-width modulation (PWM) approximates an analog control signal by switching between two voltage levels. For an ideal PWM waveform switching between 0 and V_H, its average value after suitable low-pass filtering is

where D is duty ratio between 0 and 1. PWM is efficient for motor drives, heaters, and LED control, but ripple, switching frequency, electromagnetic interference, and load dynamics must be considered. A motor driver normally requires a power stage and flyback protection rather than a microcontroller pin directly connected to the motor.

7. Switching Devices and Digital Interfaces

Digital outputs are designed for logic signals, not arbitrary power loads. A transistor or MOSFET is used when a controller must switch a relay, solenoid, lamp, fan, or motor.

For a low-side N-channel MOSFET switch, the load is connected to the positive supply and the MOSFET connects the load to ground when activated. Important checks include gate threshold versus actual drive voltage, on-resistance, drain current, voltage rating, power dissipation, switching speed, and thermal conditions. Gate threshold voltage is not the voltage at which the device is guaranteed to carry the intended current; the data sheet’s R_DS(on) specification at the available gate voltage is the relevant criterion.

Inductive loads store energy. When current is interrupted, the inductor attempts to maintain current and can generate a damaging voltage. A flyback diode placed across a DC coil provides a current path. For faster release or specialized switching, a zener clamp, TVS device, or active clamp may be selected, but the voltage and energy ratings must be verified.

Logic-level compatibility requires checking input-high and input-low thresholds, output drive current, voltage domains, rise and fall times, and whether the interface is push-pull, open-drain, or open-collector. Open-drain interfaces require a pull-up resistor and are useful for shared buses, but the pull-up value affects current and transition speed.

Common digital interfaces include UART for point-to-point serial communication, SPI for short-distance synchronous peripherals, and I²C for addressed multi-device communication. Differential standards such as RS-485 are more suitable for longer industrial cables because they reject common-mode interference, provided termination, biasing, grounding, and transceiver limits are correctly designed.

8. Protection, Grounding, and Noise Mitigation

8.1 Protection

Protection is designed from credible fault energy, not from nominal operating conditions alone. Common methods include:

• series resistors to limit fault current;
• clamp diodes or TVS devices for transient suppression;
• resettable fuses or conventional fuses for overcurrent;
• reverse-polarity protection;
• current-limited power supplies;
• galvanic isolation using optocouplers, isolated amplifiers, or isolated converters;
• shield termination and connector protection for external cables.

A protection component must withstand the expected pulse energy, repetition rate, clamping voltage, and temperature. A clamp that protects an input from overvoltage may itself fail if the upstream source can deliver substantial current.

8.2 Grounding and shielding

Ground is not automatically an ideal zero-impedance reference. Conductors have resistance and inductance, so load currents can create voltage differences between points called ground. A high-current motor return routed through a sensor ground can produce measurement error.

Star grounding can keep high-current and low-level returns separate before joining at a deliberate reference point. Ground planes reduce impedance in printed-circuit boards, but routing and return-current paths must still be considered. Differential measurement reduces sensitivity to voltage common to both inputs, while common-mode range and isolation limits remain important.

A shield reduces capacitive electric-field coupling when connected appropriately. For low-frequency analog instrumentation, shield connection at one end is sometimes used to reduce ground-loop current; at high frequencies, a 360-degree chassis termination can provide a better RF path. The correct method depends on frequency, safety requirements, equipment architecture, and applicable standards. Shielding is not a substitute for good signal routing.

8.3 Noise sources

Noise may arise from thermal resistor noise, semiconductor noise, power-supply ripple, switching edges, electromagnetic coupling, ground loops, quantization, and aliasing. Practical mitigation includes reducing bandwidth, shortening high-impedance nodes, separating power and signal routing, using differential signaling, adding local bypass capacitors, controlling edge rates when appropriate, and sampling synchronously with the process.

A bypass capacitor should be placed close to the relevant integrated-circuit supply pins, with a low-inductance return. Decoupling values and layout are important because a capacitor that is electrically distant may be ineffective at high frequencies.

9. A Systematic Interface-Design Method

A robust design process is:

1. Define the variable and required performance. Specify range, accuracy, bandwidth, response time, and fault conditions.
2. Characterize the sensor. Determine output type, excitation, source impedance, linearity, temperature coefficient, and common-mode voltage.
3. Map the signal range. Select gain, attenuation, and bias so the complete signal remains within safe input and output limits.
4. Set bandwidth and sampling. Determine the useful frequency range and add anti-aliasing filtering.
5. Check loading and drive. Confirm that the sensor, amplifier, ADC sample capacitor, and cable do not interact adversely.
6. Design protection and grounding. Consider normal operation, wiring faults, ESD, surge, and maintenance conditions.
7. Select components using guaranteed specifications. Typical values are not sufficient for safety-critical or precision requirements.
8. Validate by calculation, simulation, prototype testing, and environmental testing. Measure noise, drift, saturation, settling, and fault behavior.

Accuracy should be treated as an error budget. Contributions may include sensor error, resistor tolerance, amplifier offset, gain error, ADC quantization, reference error, temperature drift, and noise. Independent worst-case errors may be summed conservatively; statistically independent random errors may instead be combined by root-sum-square, but only when the independence assumptions are justified.

10. Illustrative Case: Load-Cell Interface for a Small Automated Press

Background

An automated benchtop press must monitor force and stop when the force reaches a programmed limit. A strain-gauge load cell has a full-scale rating of 1 kN, a sensitivity of 2 mV/V at full scale, and is excited by 5.000 V. The controller uses a 0–3.3 V, 12-bit ADC.

Problem

At full load, the bridge output is only

The signal is differential and small. It must be amplified, filtered, protected, and converted without exceeding the ADC input range.

Analysis and approach

An instrumentation amplifier with gain G = 200 would produce 2.00 V at full load. A midscale bias of 1.00 V allows measurement of positive and negative deviations, giving an ideal output range near 1.00 to 3.00 V for ±10 mV around the nominal point. However, the design must verify amplifier output swing, input common-mode range, offset, noise, and gain tolerance.

A low-pass filter with a nominal cutoff near 20 Hz is appropriate if the press-force control loop needs only slowly varying force data. For R = 8.2 kΩ and C = 1.0 µF,

The ADC sampling rate might be selected as 1 ksample/s, while the analog filter attenuates frequencies well above the force bandwidth. Digital averaging can further reduce random noise, but it cannot recover information lost through aliasing.

The bridge excitation and ADC reference may be arranged ratiometrically if the ADC architecture permits. The load-cell cable should use twisted differential conductors and an appropriate shield. The amplifier input should include carefully selected series resistance and transient clamps that do not introduce excessive offset or leakage. The controller should command the press actuator through a properly rated MOSFET, relay driver, or industrial output module; the load cell must never share a high-current return path with the actuator.

Outcome

The proposed interface uses a bridge excitation supply, instrumentation amplifier, low-pass anti-aliasing filter, protected ADC input, and separated power return. Calibration at zero and at a known reference load compensates for residual offset and gain error. A fault test verifies that disconnected, shorted, and over-range sensor conditions generate a safe stop rather than an unnoticed force command.

Learning points

The case demonstrates that a low-level sensor interface is a complete system-design problem. Gain, bandwidth, ADC range, grounding, protection, calibration, and actuator isolation must be considered together. A nominally accurate amplifier cannot compensate for poor cable routing or inadequate fault handling.

11. Limitations, Standards, and Contemporary Practice

No interface is perfectly linear, noiseless, or immune to faults. Component tolerances vary with temperature and age. Filters introduce delay. Digital conversion introduces quantization and sampling constraints. Isolation can improve safety and ground-loop performance but adds cost, propagation delay, and power requirements.

Designers should consult applicable electrical-safety, electromagnetic-compatibility, machinery-safety, and functional-safety standards. The exact standard depends on the product, installation, jurisdiction, and risk classification. Safety-related decisions require documented hazard analysis and competent review; software or AI-generated circuit suggestions must not replace measurement, simulation, data-sheet verification, or compliance testing.

Contemporary data-acquisition systems increasingly use integrated sensor interfaces, programmable-gain amplifiers, delta-sigma ADCs, isolated industrial Ethernet, and edge processing. Digital calibration and anomaly detection can identify sensor drift, intermittent wiring, or implausible combinations of measurements. Such methods are useful only when the underlying signal chain has sufficient integrity and when fail-safe behavior is defined for missing or corrupted data.

Chapter Summary

Signal conditioning connects real-world sensors to digital controllers by amplifying, filtering, shifting, isolating, protecting, and sometimes linearizing electrical signals. Voltage dividers are simple but subject to loading. Op-amps provide buffering, gain, inversion, and differential measurement, while instrumentation amplifiers are especially suitable for small bridge outputs. Bridges convert resistance changes into small differential voltages and require attention to excitation, matching, temperature, and calibration.

Filters define bandwidth and reduce noise; an analog anti-aliasing filter is required before sampling. ADC resolution is determined by reference range and bit count, but practical accuracy also depends on offset, gain, reference, linearity, noise, and acquisition settling. DACs and PWM provide analog-equivalent control signals. MOSFETs, flyback paths, logic-level checks, and suitable drivers are essential for switching loads.

Reliable interfaces also require protection, deliberate grounding, shielding, appropriate return-current paths, decoupling, and electromagnetic-compatibility thinking. A professional design begins with requirements, maps signal ranges, checks loading and bandwidth, creates an error budget, verifies fault behavior, and validates the result experimentally.

Key Terms

Signal conditioning: Electrical processing that makes a sensor signal suitable for measurement or control.

Data acquisition: Measurement, conversion, storage, and transfer of physical-process data.

Operational amplifier: High-gain differential amplifier commonly used with feedback.

Instrumentation amplifier: Precision differential amplifier with high input impedance and high common-mode rejection.

Common-mode voltage: Voltage component shared by two measurement inputs.

CMRR: Ratio of differential gain to common-mode gain.

Wheatstone bridge: Resistor network that converts resistance changes into differential voltage.

Excitation: Electrical supply applied to a sensor or bridge.

Low-pass filter: Circuit that passes low frequencies and attenuates higher frequencies.

Cutoff frequency: Frequency at which a first-order filter magnitude is 0.707 of its passband value.

Aliasing: Misrepresentation of high-frequency content as lower-frequency content after sampling.

ADC: Analog-to-digital converter.

DAC: Digital-to-analog converter.

Quantization: Mapping of continuous amplitude to discrete digital codes.

PWM: Pulse-width modulation using duty ratio to control average output.

MOSFET: Voltage-controlled semiconductor switching device.

Flyback diode: Diode that provides a current path when an inductive load is switched off.

Ground loop: Unwanted current path between nominally common reference points.

Ratiometric measurement: Measurement in which sensor excitation and conversion reference share a controlled ratio.

Anti-aliasing filter: Analog filter that limits input bandwidth before sampling.

Review Questions

Short Answer Questions

1. Why is signal conditioning usually required between a sensor and an ADC?
2. State the ideal voltage-gain equation for a non-inverting op-amp.
3. What is the purpose of a flyback diode across a relay coil?
4. Define aliasing and state one method of preventing it.
5. Why can a motor return current corrupt a low-level sensor measurement?

Long Answer Questions

1. Explain the operation of a Wheatstone bridge and discuss four sources of bridge-measurement error.
2. Compare voltage-divider, differential-amplifier, and instrumentation-amplifier interfaces.
3. Explain ADC resolution, quantization, reference uncertainty, and acquisition-time limitations.
4. Develop a practical grounding, shielding, and decoupling strategy for a motorized mechatronic system.
5. Describe a complete design procedure for conditioning a low-level analog sensor for a microcontroller.

Analytical / Application Questions

1. A 0–25 mV sensor must be mapped to a 0.2–3.0 V ADC input. Determine a suitable linear gain and bias arrangement, then identify practical headroom and tolerance checks.
2. Design a first-order RC anti-aliasing filter for a signal bandwidth of 10 Hz and discuss why selecting the cutoff exactly at 10 Hz may be inadequate.
3. A load-cell amplifier output becomes unstable when a motor starts. Analyze possible coupling, grounding, filtering, shielding, and supply-related causes, and propose a test plan.

Multiple Choice Questions with Answers

Question 1

A 12-bit ideal ADC with a 0–3.3 V reference has an approximate LSB size of:

A. 0.806 mV

B. 3.3 mV

C. 12.1 mV

D. 0.0806 mV

Correct Answer: A. 0.806 mV

Question 2

For a first-order RC low-pass filter, the −3 dB cutoff frequency is:

A. RC/(2π)

B. 1/(2πRC)

C. 2πRC

D. 1/(RC²)

Correct Answer: B. 1/(2πRC)

Question 3

The principal purpose of an instrumentation amplifier in a bridge-sensor interface is to provide:

A. High differential gain and high common-mode rejection

B. Power switching with low drain resistance

C. Digital isolation without any power supply

D. Frequency generation for the sensor excitation

Correct Answer: A. High differential gain and high common-mode rejection

Question 4

A flyback diode is used with a DC relay coil primarily to:

A. Increase the coil resistance during normal operation

B. Provide a path for inductive current when switching off

C. Convert the coil current into a digital signal

D. Increase the controller output voltage

Correct Answer: B. Provide a path for inductive current when switching off

Question 5

If a PWM signal switches between 0 V and 5 V with duty ratio 0.40, its ideal filtered average is:

A. 0.5 V

B. 2.0 V

C. 4.0 V

D. 12.5 V

Correct Answer: B. 2.0 V

Question 6

Which practice most directly reduces aliasing before an ADC?

A. Adding an analog low-pass filter before sampling

B. Increasing the sensor cable length

C. Removing all ADC reference capacitors

D. Connecting a motor return through the sensor ground

Correct Answer: A. Adding an analog low-pass filter before sampling

Question 7

In an ideal non-inverting op-amp, R_f = 9 kΩ and R_g = 1 kΩ. The voltage gain is:

A. 0.10

B. 9

C. 10

D. −9

Correct Answer: C. 10

Question 8

Why is the MOSFET data-sheet R_DS(on) specification more useful than threshold voltage for switch selection?

A. Threshold voltage guarantees the desired load current

B. R_DS(on) is specified for a particular gate-drive condition and conduction performance

C. R_DS(on) determines the sensor bridge sensitivity

D. Threshold voltage is independent of temperature and current

Correct Answer: B. R_DS(on) is specified for a particular gate-drive condition and conduction performance

Question 9

Which interface characteristic is especially useful for rejecting noise coupled equally into two conductors?

A. Single-ended measurement with a long ground lead

B. Differential signaling with adequate common-mode range

C. A larger unfiltered sensor impedance

D. A floating ADC with no reference

Correct Answer: B. Differential signaling with adequate common-mode range

Question 10

A quarter-bridge with a small resistance change has an output magnitude approximately equal to:

A. V_ex∆R/R

B. 4V_ex∆R/R

C. V_ex∆R/(4R)

D. R/(4V_ex∆R)

Correct Answer: C. V_ex∆R/(4R)