2.2 Electronics & Instrumentation

Brain potentials reaching the scalp are tiny — on the order of 10 to 100 microvolts (uV) — and they are buried in a sea of larger environmental signals. EEG instrumentation exists to amplify the wanted signal, reject the unwanted noise, and present a calibrated, faithful display. The R.EEG.T. exam expects you to know not just which knob to turn but why the underlying electronics behave as they do.

The Differential Amplifier

Every EEG channel is built around a differential amplifier. It has two inputs (input 1 and input 2) and amplifies the difference between them while rejecting whatever is common to both:

• Output is proportional to (Input 1 - Input 2).
• A signal that appears identically on both inputs (a common-mode signal) is theoretically cancelled to zero.

This is the single most important concept in EEG electronics. Environmental noise — most importantly 60 Hz power-line interference — couples almost equally onto both electrodes of a pair. Because it is common to both inputs, the differential amplifier subtracts it away while preserving the brain signal, which differs between the two scalp sites.

Common Mode Rejection Ratio (CMRR)

No amplifier is perfect, so we quantify how well it rejects common-mode signals with the Common Mode Rejection Ratio (CMRR). It is the ratio of the differential gain to the common-mode gain, expressed in decibels (dB).

The key relationship to memorize is the dB-to-factor conversion, because exam questions test it directly:

Every 20 dB equals a tenfold (10x) increase in rejection. A modern EEG amplifier typically has a CMRR of at least 100 dB (100,000:1) or better. Higher CMRR is better — it means more aggressive cancellation of shared 60 Hz and other environmental noise. A practical corollary: balanced electrode impedances help the amplifier 'see' the noise equally on both inputs, so keeping impedances low and similar maximizes the effective common-mode rejection.

Filters: Shaping the Recorded Bandwidth

Filters discard frequencies you do not want so that the EEG of interest stands out. The exam uses both EEG-historical names and the engineering names — know both.

A frequent exam trap: the low-frequency filter passes high frequencies (it is a high-pass filter), and the high-frequency filter passes low frequencies (a low-pass filter). The filter is named for what it attenuates, while the engineering name describes what it passes.

Using Filters Wisely

Raising the LFF (for example from 1 Hz to 5 Hz) reduces slow sweat and movement artifact but will also distort or remove legitimate delta and theta activity, so it must be used cautiously. Lowering the HFF (for example from 70 Hz to 35 Hz) reduces muscle (EMG) artifact but can blunt true spikes and fast activity. The notch filter should be a last resort: it removes 60 Hz noise but also attenuates real EEG near 60 Hz and can hide that the true problem is high or unbalanced electrode impedance. Always fix the electrode problem first.

Sensitivity, Gain, and Calibration

Sensitivity controls how large the displayed signal is. It is expressed in microvolts per millimeter (uV/mm) for paper-style display or microvolts per the equivalent screen unit in digital systems.

• Sensitivity is an inverse scale: a setting of 5 uV/mm produces a larger deflection than 10 uV/mm for the same input.
• Lower the sensitivity number to make low-amplitude activity more visible; raise it when high-amplitude activity goes off scale.

Gain is the raw amplification factor of the channel (output voltage divided by input voltage). Sensitivity is the practical, clinician-facing way of expressing the net effect of gain on the display.

Calibration

Before and after a recording the technologist runs a calibration. A known square-wave voltage (commonly 50 uV) is fed to every channel so the deflection can be verified equal across channels and correctly sized for the chosen sensitivity. The square wave also reveals the time constant: the slope of its decay confirms the low-frequency filter, while the sharpness of its rising edge confirms the high-frequency filter. A biological ('bio') calibration then feeds the same electrode pair to all channels so every channel can be confirmed to reproduce identical EEG, isolating channel-specific faults.

Digital EEG: Sampling, Nyquist, and Aliasing

Modern EEG is digital. An analog-to-digital (A/D) converter measures the continuous analog voltage at fixed time intervals and stores each measurement as a number — this is sampling. Two specifications govern fidelity:

• Sampling rate (Hz): how many samples per second are taken per channel.
• A/D resolution (bits): how finely each sample's voltage is quantized (more bits = finer amplitude steps and a wider dynamic range).

The Nyquist Theorem

The Nyquist (sampling) theorem states that the sampling rate must be at least twice the highest frequency you want to record accurately. The highest faithfully recordable frequency is called the Nyquist frequency and equals half the sampling rate.

Aliasing

If a signal contains frequencies above the Nyquist frequency and is not removed first, those fast components are misrepresented as false low frequencies — a distortion called aliasing. Aliasing is prevented by an anti-aliasing filter: an analog low-pass filter applied before the A/D converter to remove components above the Nyquist frequency, plus sampling well above twice the highest frequency of interest. This is exactly why routine digital EEG samples at 256 Hz or higher even though clinical EEG of interest rarely exceeds 70 Hz.