Why do ECG, EEG, and EMG all require an instrumentation amplifier rather than a simple op-amp gain stage?

All three biosignals share two demanding characteristics that a simple single op-amp stage cannot adequately handle: they are extremely small (microvolts for EEG, tens of microvolts to low millivolts for EMG, roughly 0.5-4 mV for ECG) riding on top of much larger common-mode voltages and interference (often from 60/50 Hz power-line coupling into the body and cabling), and they are measured differentially between two or more electrodes with high, variable, and often asymmetric electrode-skin impedance. An instrumentation amplifier's three-op-amp topology provides very high input impedance (avoiding loading and further imbalancing already-high and asymmetric electrode impedances), high common-mode rejection ratio (rejecting the large shared interference signal while amplifying only the tiny differential biosignal), and precisely settable gain via a single resistor — a combination of properties a basic differential or single-ended op-amp gain stage cannot reliably deliver at the input impedance and CMRR levels biopotential measurement requires.

What is a right-leg drive circuit and why does ECG equipment use it instead of just grounding the patient?

A right-leg drive (RLD) circuit actively senses the common-mode voltage present on the patient's body (via the differential amplifier's own internal common-mode sensing) and drives that same common-mode signal back into the patient, inverted, through a dedicated electrode (traditionally placed on the right leg) — creating a negative feedback loop that actively cancels common-mode interference rather than passively grounding the patient. This active approach achieves substantially better common-mode noise rejection than simple grounding could alone, and critically, it is designed with a current-limiting resistor specifically so that even under a single-fault condition (such as the drive amplifier saturating to a supply rail), the current that could flow into the patient through that electrode remains far below any level capable of causing physiological harm — an essential IEC 60601 patient-safety consideration for any circuit that actively drives current into a patient.

Why is EEG measurement so much more demanding than ECG in terms of amplifier noise performance?

EEG signals are roughly two to three orders of magnitude smaller than ECG signals — typically 10-100 microvolts for EEG compared to 0.5-4 millivolts for ECG — while both signals ride on top of comparable levels of environmental and physiological interference (power-line coupling, muscle artifact, electrode noise). This means the required signal-to-noise ratio and input-referred amplifier noise specification for a usable EEG system is dramatically more demanding than for ECG: an amplifier noise floor that would be completely negligible relative to a millivolt-scale ECG signal can be comparable in size to a tens-of-microvolt EEG signal, actively corrupting the measurement. This is why dedicated EEG amplifier front ends require ultra-low-noise instrumentation amplifiers, extremely high input impedance (often gigaohms, to avoid attenuating the signal against typically higher and more variable EEG electrode-skin impedance than ECG electrodes present), and often active shielding or driven-shield cabling techniques that simpler ECG front ends can sometimes forgo.

Engineering·11 min read·August 7, 2026

📈 Biosignal Acquisition Fundamentals: ECG, EEG, EMG, Amplifier Design, and Noise Rejection

An engineering guide to biosignal acquisition — the electrical origin of ECG, EEG, and EMG signals, instrumentation amplifier design for biopotential measurement, and the noise and artifact sources every biomedical instrumentation design must reject.

By EngineersUniverse Editorial Team

Engineering Editorial & Technical Review Team

Published August 7, 2026

The Body Generates Its Own Measurable Electrical Signals

Biosignal acquisition is the discipline of detecting, amplifying, filtering, and digitizing the small electrical signals the human body naturally generates — signals that originate from the electrochemical activity of excitable cells (neurons and muscle fibers) and can be measured non-invasively at the skin surface through electrodes. The three most clinically important biopotential signals are the electrocardiogram (ECG, cardiac electrical activity), electroencephalogram (EEG, brain electrical activity), and electromyogram (EMG, skeletal muscle electrical activity) — and while each has a distinct physiological origin and clinical use, the underlying instrumentation engineering challenge they present is remarkably consistent: extract a very small, physiologically meaningful signal from a noisy, high-impedance, and safety-constrained measurement environment.

The Electrical Origin of Each Signal

ECG — Cardiac Electrical Activity

The heart's coordinated contraction is triggered by an electrical depolarization wave that originates at the sinoatrial node and propagates through the atria and ventricles in a specific, repeatable sequence. That propagating wavefront of depolarizing cardiac tissue generates a time-varying electrical dipole whose projection onto the body surface produces the characteristic ECG waveform — the P wave (atrial depolarization), QRS complex (ventricular depolarization, the largest and most clinically prominent feature), and T wave (ventricular repolarization). ECG signal amplitude at the skin surface typically ranges from roughly 0.5 to 4 millivolts, and the clinically relevant frequency content spans roughly 0.05 Hz to 150 Hz (with diagnostic-quality ECG requiring the lower end of that range to accurately capture ST-segment features, while monitoring-grade ECG can use a narrower bandwidth).

EEG — Brain Electrical Activity

EEG measures the summed postsynaptic electrical activity of large populations of cortical neurons firing in loose synchrony, detected as small voltage fluctuations at the scalp. Individual neuron action potentials are far too small and too localized to detect at the scalp; what EEG actually measures is the aggregate field potential from thousands to millions of neurons whose activity happens to be sufficiently synchronized and geometrically aligned to produce a summed signal large enough to reach the scalp surface, typically in the 10-100 microvolt range — far smaller than ECG, as detailed in this article's FAQ. EEG frequency content is clinically divided into named bands (delta, theta, alpha, beta, gamma, roughly spanning 0.5-100 Hz) that correlate with different brain states and are central to clinical EEG interpretation.

EMG — Skeletal Muscle Electrical Activity

EMG measures the electrical activity generated when a motor neuron triggers a muscle fiber to contract — the motor unit action potential. Surface EMG (measured non-invasively at the skin over a muscle) detects the summed activity of many motor units beneath the electrode, with amplitude typically ranging from tens of microvolts to a few millivolts depending on contraction strength and electrode placement, and frequency content roughly spanning 10-500 Hz. Surface EMG is used clinically for neuromuscular diagnosis and increasingly in prosthetics and rehabilitation engineering as a control signal — a myoelectric prosthetic hand, for example, uses surface EMG from residual-limb muscles as its primary user control input.

Instrumentation Amplifier Design for Biopotential Measurement

As detailed in this article's FAQ, all three biosignal types require an instrumentation amplifier front end rather than a simple op-amp gain stage, because they share the combination of very small signal amplitude, high and variable source (electrode-skin) impedance, and large common-mode interference. The classic three-op-amp instrumentation amplifier topology — two input buffer stages providing high input impedance and a single-resistor-settable first-stage gain, followed by a difference amplifier stage rejecting the common-mode component — is the standard building block, though modern biomedical instrumentation increasingly uses integrated instrumentation amplifier or dedicated biopotential front-end ICs that combine this function with additional features like right-leg-drive support, programmable gain, and built-in anti-aliasing filtering.

Noise and Artifact Sources

Real-world biosignal acquisition must actively reject several characteristic noise and artifact sources, each requiring a specific engineering countermeasure:

Power-line interference (50/60 Hz) — capacitively and inductively coupled from nearby AC wiring and equipment into the patient's body and lead cables; addressed primarily through high common-mode rejection ratio (CMRR) in the front-end instrumentation amplifier, active right-leg-drive circuitry (for ECG/EEG), and shielded, twisted-pair lead cabling.
Motion artifact — voltage transients generated at the electrode-skin interface itself when mechanical movement disturbs the electrochemical double-layer that forms between the electrode and skin; addressed through electrode design (Ag/AgCl electrodes with conductive gel are far more motion-stable than dry metal contacts), secure mechanical electrode fixation, and, in some systems, adaptive signal processing that detects and flags motion-corrupted segments.
Muscle artifact (EMG contamination of ECG/EEG) — since all three biosignals share overlapping frequency content and are all detectable at the skin surface, unwanted muscle activity near a recording site can contaminate an ECG or EEG recording; addressed partly through electrode placement strategy and partly through frequency-domain and, increasingly, algorithmic artifact-rejection techniques.
Baseline wander — slow drift in the signal baseline caused by respiration, electrode impedance changes, or DC offset drift; addressed through appropriate high-pass filtering (careful not to set the cutoff so high that it distorts genuine low-frequency signal content, particularly the ECG's diagnostically important ST segment).

From Analog Front End to Digital Signal Processing

After analog amplification, filtering (typically a bandpass configuration matched to the specific signal's clinically relevant frequency range, plus an anti-aliasing low-pass filter set below half the analog-to-digital converter's sampling rate per the Nyquist criterion), and digitization, modern biosignal systems apply digital signal processing for further noise reduction, feature extraction (such as automated QRS detection in ECG), and, in advanced systems, real-time algorithmic interpretation. The fundamental engineering discipline connecting the analog front end and the digital back end remains the same throughout: preserve the small, clinically meaningful signal while rejecting everything else competing for the same frequency space and electrode contact.

Topics covered

biosignal acquisitionECG amplifier designEEG signal processingEMG measurementinstrumentation amplifier biopotentialcommon mode rejection ratio biosignalright leg drive circuitmotion artifact rejectionbiopotential electrode