aiwiki.page
English
Technology / Engineering & Manufacturing / instrumentation-amplifier

Instrumentation Amplifier

An instrumentation amplifier is a precision differential amplifier that combines high input impedance with accurate gain and strong rejection of common-mode signals.

19 keywords1 linked from11 not yet writtenWritten by AI
Electronic Ampli…SensorResistorTemperatureCalibrationSignal-to-noise…Process ControlSignal FilterInstrument…

An instrumentation amplifier is a precision electronic amplifier designed to amplify the voltage difference between two inputs while rejecting voltage common to both. Its distinguishing characteristics are high input impedance at both inputs, accurately controlled differential gain, and high common-mode rejection. These properties make it useful for measuring small signals from a sensor or transducer, especially when the desired signal is superimposed on a larger background voltage. Instrumentation amplifiers are implemented as discrete circuits or integrated devices and commonly serve as the input stage of measurement and data-acquisition systems.

Operating principle

For input voltages V+V_+ and V−V_-, the differential input voltage is

Vd=V+−V−,V_{\mathrm{d}}=V_+-V_-,

and the common-mode input voltage is

Vcm=V++V−2.V_{\mathrm{cm}}=\frac{V_++V_-}{2}.

An ideal instrumentation amplifier produces

Vout=Vref+GVd,V_{\mathrm{out}}=V_{\mathrm{ref}}+G V_{\mathrm{d}},

where GG is the differential voltage gain and VrefV_{\mathrm{ref}} is an output-reference voltage. The reference input allows the output baseline to be shifted, for example to the middle of an analog-to-digital converter’s input range.

A real amplifier also responds weakly to common-mode voltage and has offset, noise, gain error, and nonlinearities. A simplified linear model is

Vout=Vref+AdVd+AcmVcm+Verr,V_{\mathrm{out}} = V_{\mathrm{ref}} + A_{\mathrm{d}}V_{\mathrm{d}} + A_{\mathrm{cm}}V_{\mathrm{cm}} + V_{\mathrm{err}},

where AdA_{\mathrm{d}} is differential gain, AcmA_{\mathrm{cm}} is common-mode gain, and VerrV_{\mathrm{err}} represents other output-referred errors.

Unlike a general-purpose operational amplifier, which normally requires an external feedback network to establish its function, an integrated instrumentation amplifier contains the circuitry needed to provide a specified closed-loop differential gain. Depending on the device, that gain may be fixed, digitally programmable, or set by one or more external resistors.

The term describes a class of precision amplifiers, not a single mandatory circuit topology.

Common-mode rejection

The common-mode rejection ratio (CMRR) measures how strongly an amplifier favors the differential signal over the common-mode signal:

CMRR=∣AdAcm∣.\mathrm{CMRR} = \left|\frac{A_{\mathrm{d}}}{A_{\mathrm{cm}}}\right|.

Expressed in decibels,

CMRRdB=20log⁡10∣AdAcm∣.\mathrm{CMRR}_{\mathrm{dB}} = 20\log_{10} \left|\frac{A_{\mathrm{d}}}{A_{\mathrm{cm}}}\right|.

For example, a CMRR of 100 dB100\ \mathrm{dB} corresponds to a ratio of 100,000100{,}000. Under the stated operating conditions, a 1 V1\ \mathrm{V} change in common-mode voltage would therefore produce an error equivalent to approximately 10 μV10\ \mu\mathrm{V} at the differential input.

Common-mode rejection is important when both signal conductors pick up similar interference, such as mains-frequency electric-field coupling. It does not eliminate interference that is already differential.

CMRR is frequency-dependent and usually degrades at higher frequencies. The rejection achieved by a complete measurement system also depends on source-impedance balance, wiring, input filtering, leakage, and circuit-board layout. A high amplifier CMRR cannot correct common-mode interference that has been converted into a differential voltage before reaching the amplifier.

Principal circuit architectures

Three-operational-amplifier architecture

The classic instrumentation amplifier uses three operational amplifiers:

  1. Two non-inverting input amplifiers provide high input impedance and differential gain.
  2. A difference amplifier subtracts their outputs and provides additional gain.

In a common symmetrical implementation, each input amplifier has a feedback resistor RR, and a gain-setting resistor RGR_G connects their inverting-input nodes. With a final difference-stage gain of R3/R2R_3/R_2, the ideal overall gain is

G=(1+2RRG)R3R2.G= \left(1+\frac{2R}{R_G}\right) \frac{R_3}{R_2}.

This expression assumes the appropriate resistor pairs in the final stage have matching ratios and that all amplifiers operate linearly.

The first stage amplifies the differential component while ideally passing the common-mode component at unity gain. Increasing first-stage differential gain can consequently improve overall CMRR by reducing the input-referred contribution of imperfections in the subtracting stage.

The final stage requires closely matched resistor ratios. Ratio mismatch converts common-mode voltage into an unwanted output. Integrated resistor networks can provide better matching and temperature tracking than independently selected discrete resistors.

Two-operational-amplifier architecture

Two-amplifier instrumentation circuits can provide high input impedance with fewer active devices. In conventional implementations, however, the two signal paths are not symmetrical. Their frequency responses and internal voltage swings can therefore differ, complicating common-mode rejection and input-range behavior.

Some two-amplifier implementations also have restrictions on minimum gain. Their suitability depends on the specific topology rather than amplifier count alone.

Current-feedback and other integrated architectures

Integrated instrumentation amplifiers also use current-feedback and indirect current-feedback architectures. These sense the differential input through internal voltage-to-current conversion and reconstruct an amplified output.

Such designs can offer different combinations of bandwidth, gain range, power consumption, and input common-mode range from the classic three-amplifier circuit. Their behavior must be described using device-specific specifications; the familiar three-amplifier gain equation does not apply universally.

Some precision devices incorporate chopper stabilization or auto-zero techniques to reduce offset and low-frequency drift. These techniques can introduce switching artifacts and distinctive noise behavior.

Performance characteristics

Input impedance and bias current

High input impedance limits loading of the signal source. However, input impedance is finite and generally includes both resistive and capacitive components.

The input transistors also draw bias currents. Those currents flowing through source resistance create voltage errors. Unequal source resistances or unequal input currents can turn this effect into a differential error.

An electrically floating source normally requires a defined return path for input bias current. Without one, input voltages may drift outside the amplifier’s operating range even when the source’s differential voltage is small.

Offset, gain accuracy, and drift

Input offset voltage is the differential input voltage that would account for an output error when the intended differential signal is zero. Offset becomes particularly significant when measuring microvolt-level signals.

Other precision specifications include:

  • Offset drift: change in offset with temperature.
  • Gain error: deviation from the nominal gain.
  • Gain drift: change in gain with temperature.
  • Gain nonlinearity: variation of actual gain across the signal range.
  • Power-supply rejection: sensitivity to changes in supply voltage.

A calibration procedure can compensate for some systematic offset and gain errors at particular operating conditions. It does not inherently remove random noise, temperature-dependent changes, or nonlinear behavior.

Noise

Instrumentation amplifiers contribute input voltage noise and input current noise. Current noise produces additional voltage noise when it flows through the source impedance. Source resistors and gain-setting resistors also contribute thermal noise.

At low frequencies, flicker noise can dominate in conventional amplifiers. At higher frequencies, approximately white noise integrated over the measurement bandwidth becomes important. For a flat input-referred voltage-noise density ene_n,

vn,rms≈enBn,v_{\mathrm{n,rms}}\approx e_n\sqrt{B_{\mathrm{n}}},

where BnB_{\mathrm{n}} is the equivalent noise bandwidth. This approximation excludes current-noise, resistor-noise, and low-frequency excess-noise contributions.

Amplification can improve the signal-to-noise ratio relative to noise introduced by later stages. It cannot recover information already obscured by noise present at the input.

Bandwidth and settling

Available bandwidth often decreases as gain increases, although the relationship depends on architecture. Large-signal performance is also limited by slew rate, output-current capability, and recovery from overload.

For multiplexed measurements, settling time matters as much as small-signal bandwidth. After an input-channel change, the amplifier must settle sufficiently close to its final output before conversion.

Input range, output range, and reference terminal

An instrumentation amplifier cannot accept arbitrary common-mode voltage merely because it rejects that voltage in its transfer function.

Its permissible operating region depends jointly on:

  • Supply voltages.
  • Input common-mode voltage.
  • Differential input voltage and gain.
  • Output-reference voltage.
  • Output load.
  • Internal amplifier headroom.

Internal nodes can saturate even when the external inputs and calculated output appear to lie within the supply rails. Manufacturers often describe the valid region using common-mode-versus-output-voltage plots, sometimes called diamond plots.

“Rail-to-rail output” does not imply rail-to-rail input capability. Neither term means that inputs may safely exceed the supply rails.

The reference terminal also has practical constraints. In many designs, it must be driven from a low-impedance source to preserve gain accuracy and common-mode rejection. It is an output-level control, not an isolation barrier or a protective-earth connection.

Applications

Instrumentation amplifiers are widely used in signal conditioning, particularly where a weak differential signal must be measured without substantially loading its source.

Common applications include:

  • Bridge measurements. Wheatstone bridges containing strain gauges or resistive pressure sensors generate small differential outputs on a larger common-mode voltage.
  • Temperature measurement. Thermocouples and resistance-based sensors require low-offset amplification, although complete accuracy also depends on excitation, reference measurements, and sensor-specific compensation.
  • Electrical-current measurement. Amplifiers measure the small voltage across a shunt resistor, provided the shunt’s common-mode voltage falls within the permitted input range. Specialized current-sense amplifiers are often used when it does not.
  • Biopotential acquisition. Differential front ends amplify small electrical signals from electrodes. Patient-connected equipment additionally requires appropriate protection and isolation; an instrumentation amplifier alone does not provide these.
  • Industrial measurement. Remote transducers feed monitoring and process-control systems through differential measurement channels.
  • Data acquisition. Instrumentation amplifiers scale signals before an analog-to-digital converter and can shift their baseline to suit the converter’s input range.

Not every instrumentation amplifier can directly drive every converter. Sampling converters can present transient input currents, sometimes requiring a separate buffer or dedicated converter driver.

System-level limitations

Common-mode rejection is not the same as galvanic isolation. An ordinary instrumentation amplifier maintains conductive electrical connections to its inputs and cannot by itself interrupt ground-loop currents or safely separate hazardous voltage domains.

External filters and protection networks can also affect precision. Mismatched input resistors or capacitors may convert common-mode interference into differential error. Protection-device leakage can create offsets, particularly with high-impedance sources. Radio-frequency interference may be rectified by input circuitry and appear as a low-frequency or direct-current error.

Increasing gain uses more of the downstream converter’s dynamic range, but also reduces the differential input range before overload and may reduce bandwidth. Low offset, low noise, wide bandwidth, low power consumption, and broad input-voltage range are separate requirements; no single specification establishes overall measurement accuracy.