Zero-Drift Amplifiers: A Keystone To The Elusive Ideal Amp?

Oct. 8, 2010
$jq().ready( function() \{ setupSidebarImageList(); \} );

Auto-zero amplifier

Auto-zero phase

ADA4051 chopper amplifier


Amplifier topologies

Output ripple

Circuit gain

Intermodulation distortion

Improve noise

Amplifier Design

Zero-drift amplifiers dynamically correct their offset voltage and reshape noise density. Two commonly used types—auto-zero amplifiers and choppers—achieve nanovolt-level offsets and extremely low offset drifts due to time and temperature. The amplifier’s 1/f noise is also seen as a dc error, so it’s removed as well.

Designers derive many benefits from zero-drift amplifiers, considering nuisances such as temperature drift and 1/f noise are otherwise very difficult to eliminate. In addition, zero-drift amplifiers offer higher open-loop gain, power-supply rejection, and common-mode rejection than standard amplifiers. On top of that, their overall output error is less than that obtained by a standard precision amplifier in the same configuration.

Zero-drift amplifiers are used in systems with an expected design life of greater than 10 years, as well as in signal chains that use high closed-loop gains (>100) with low-frequency (<100 Hz), low-amplitude level signals. Examples can be found in precision weigh scales; medical instrumentation; precision metrology equipment; and infrared, bridge, and thermopile sensor interfaces.

Auto-Zero Technique
Auto-zero amplifiers usually correct for input offset in two clock phases. During clock phase A, the switches that are labeled φA are closed, while switches labeled φB are open, as seen in a diagram of an AD8571 (Fig. 1). The offset voltage of the nulling amplifier is measured and stored on capacitor CM1.

During clock phase B (Fig. 2), switches labeled φB are closed, while switches labeled φA are open. The offset voltage of the main amplifier is measured and stored on capacitor CM2, while the stored voltage on capacitor CM1 adjusts for the offset of the nulling amplifier. The overall offset is then applied to the main amplifier while processing the input signal.

The sample-and-hold function turns auto-zero amplifiers into sampled-data systems, making them prone to aliasing and fold-back effects. At low frequencies, noise changes slowly, so the subtraction of two consecutive noise samples results in true cancellation. This correlation diminishes at higher frequencies, with subtraction errors causing wideband components to fold back into the baseband. Thus, auto-zero amplifiers have more in-band noise than standard op amps.

To reduce low-frequency noise, the sampling frequency has to be increased, but this introduces additional charge injection. The signal path includes only the main amplifier, so relatively large unity-gain bandwidth can be obtained.

Chopping Technique
The ADA4051 chopper amplifier employs a local auto-correction feedback (ACFB) loop (Fig. 3). The main signal path includes input chopping network CHOP1, transconductance amplifier GM1, output chopping network CHOP2, and transconductance amplifier GM2.

CHOP1 and CHOP2 modulate the initial offset and 1/f noise from GM1 up to the chopping frequency. Transconductance amplifier GM3 senses the modulated ripple at the output of CHOP2. Chopping network CHOP3 demodulates the ripple back to dc. All three chopping networks switch at 40 kHz. Finally, transconductance amplifier GM4 nulls the dc component at the output of GM1, which would otherwise appear as ripple in the overall output.

Continue on next page

The switched capacitor notch filter (SCNF) selectively suppresses the undesired offset-related ripple without disturbing the desired input signal from the overall output. It’s synchronized with the chopping clock to perfectly filter out the modulated components.

In addition, the two techniques (Fig. 4) can be combined. The AD8628 zero-drift amplifier uses both auto-zeroing and chopping to reduce the energy at the chopping frequency, while keeping the noise very low at lower frequencies. This combined technique allows for wider bandwidth than is possible with conventional zero-drift amplifiers.

Auto-Zeroing Vs. Chopping
Auto-zeroing uses sampling to correct offset, while chopping uses modulation and demodulation. Because sampling causes noise to fold back into baseband, auto-zero amplifiers have more in-band noise. To suppress noise, more current is used. Therefore, the devices typically dissipate more power.

Choppers have low-frequency noise consistent with their flat-band noise, but produce a large amount of energy at the chopping frequency and its harmonics. Output filtering may be required, so these amplifiers are most suitable in low-frequency applications. Comparing typical noise characteristics helps illustrate the differences between auto-zero and chopping techniques (Fig. 5).

Application Suggestions
Zero-drift amplifiers are composite amplifiers that use digital circuitry to dynamically correct for analog offset errors. The design techniques discussed so far greatly improve a number of op-amp parameters—offset voltage, offset drift with time and temperature, and the shape of the noise curve—at the expense of a few others. This tradeoff, common in amplifier design, is one reason why there’s no claim of the “ideal amplifier.”

When using zero-drift amplifiers, application issues that must be considered include charge injection, clock feedthrough, intermodulation distortion, and overload recovery time. Charge injection, due to the switching action of choppers and auto-zero amplifiers, will appear at the amplifier inputs. For example, if a zero-drift amplifier is configured in a non-inverting configuration, small ripples will appear on the output due to input switching action (Fig. 6).

The magnitude of the charge injection is independent of temperature, but will grow if there’s an increase in circuit gain, source resistance, or the gain setting resistor. This is evident when examining the error voltage due to charge injection versus source resistance (Fig. 7).

Application-level solutions can easily reduce the contribution of this error. There are several ways to reduce the effects of charge injection:
• Add a capacitor in the feedback to limit the signal bandwidth.
• Use lower source and feedback resistors.
• Build an active or passive filter after the amplification stage.
• Add a resistor equal to the parallel combination of RF and RS to the non-inverting input to help cancel the IB effect.
• Use a device, such as the ADA4051, that has ACFB.

Clock feedthrough may occur if the amplifier isn’t designed well or if it uses a pure chopping technique. Figure 8 shows the artifacts of the internal clock over the frequency spectrum.

Signals with frequencies greater than the auto-zero frequency can be amplified. An auto-zeroed amplifier’s speed is contingent on the gain-bandwidth product, which depends on the main amplifier, not the nulling amplifier. The auto-zero frequency indicates when switching artifacts will start to occur.

Continue on next page

As the input approaches the chopping or auto-zero frequency, intermodulation distortion (IMD) is introduced, with larger errors occurring as the input frequency gets closer to the clock frequency (Fig. 8). IMD between the high-frequency input signal and the chopping frequency creates tones at frequencies fCHOP – fIN and fCHOP + fIN. By applying clever design techniques and a combination of chopping and auto-zeroing schemes, the intermodulation distortion of the AD8628 is reduced by 12 dB compared to the AD8551 family.

IMD can be improved by using a pseudorandom auto-zero frequency. The AD8571 family, for example, uses a clock that varies between 2 kHz and 4 kHz. Another possibility is to add filters around the amplifier, where the clock noise (Fig. 9) is shown after filtering. Selecting the right cutoff frequency can enhance circuit response.

The overload recovery time of zero-drift amplifiers is typically longer than that of standard CMOS amplifiers. If the inputs of an auto-zero amplifier are separated by a large amount for any reason, the output will saturate. The nulling amplifier treats this as an offset and tries to null the error. This sends the main amplifier further into saturation and prolongs the recovery time.

Amplifiers with built-in intelligence that recognizes overload (e.g., the AD8628) make it possible recover from overload in as little as 30 µs. Devices without this technology can take as long as 40 ms to recover.

Selecting A Zero-Drift Amplifier
All zero-drift amplifiers offer:

• Low offset voltage ( <10 µV max) over the entire VCM • Super-low offset voltage drift (<40 nV/°C) over time and temperature • Reshaped noise that eliminates 1/f noise • Very high open-loop gain, common-mode rejection ratio (CMRR), and power-supply rejection ratio (PSRR) • High input impedance • Extreme temperature operation (up to 200°C) • No need for external trimmingChoppers are a good choice for low-power, low-frequency applications (<100 Hz), while auto-zero amplifiers are better for wideband applications. For applications that require low noise, no switching glitch, and wide bandwidth, the optimal solution is an amplifier that combines auto-zero and chopping techniques. Examining the design tradeoffs will help designers make the proper decision (see the table).

1. “Bridge-Type Sensor Measurements Are Enhanced by Auto-Zeroed Instrumentation Amplifiers,”
2. “MT-055: Chopper Stabilized (Auto-Zero) Precision Op Amps,”
3. “Demystifying Auto-Zero Amplifiers-Part 1,
4. “Demystifying Auto-Zero Amplifiers-Part 2,”


To join the conversation, and become an exclusive member of Electronic Design, create an account today!