Zero-drift amplifiers have very low input offset voltage (Vos), low offset drift, and no flicker noise. The two main types of zero-drift amplifiers are chopper and auto-zero amplifiers. This paper focuses on chopper amplifiers because these amplifiers are by far the most common type of zero-drift amplifier and most of the chopper concepts apply to auto-zero devices as well. Chopper amplifiers achieve the low Vos and Vos drift through the use of an internal calibration circuit that uses metal-oxide semiconductor field-effect transistor (MOSFET) switches to commutate the inputs. However, this calibration technique generates current transients within the amplifiers input bias current. These transients flow through the amplifiers feedback network and source impedance generating additional offset voltage, offset drift, and transient noise tones. The magnitude of this error increases when large feedback resistances or source impedances are used. Furthermore, this error is dependent on the details of the chopper amplifier design, so that some products are more susceptible to Vos shift than others. This white paper provides details on how the magnitude of offset, drift, and noise can be impacted by source and feedback impedance for different chopper amplifier products. The document covers a method for selecting the best chopper amplifier for your application, and clarifies when a chopper amplifier is not an appropriate choice.
Zero-drift amplifiers use an internal calibration method to minimize the amplifiers input offset voltage (Vos). Since this calibration happens continuously, the change in offset over temperature is also minimized. Zero-drift amplifiers additionally improve some other parameters that relate to how offset changes versus system or environmental factors. For example, power supply rejection ratio (PSRR) is a measurement of how the amplifiers offset is impacted by changing in power supply voltage, so this specification is generally better for zero-drift amplifiers than traditional topologies. Power supply rejection ratio (PSRR), common-mode rejection ratio (CMRR), and open loop gain (AOL) are all measurements of how the Vos is impacted by changing different amplifier operating conditions, so those specifications are generally much better for zero-drift amplifiers than for traditional topologies. Similarly, electromagnetic rejection ratio (EMIRR) is a measurement of how offset changes versus applied electromagnetic interference, so this is also generally improved in choppers.
For traditional amplifiers the noise increases at low frequency (called 1/f or flicker noise). Flicker noise can be thought of as a variation of input offset voltage versus time. Thus, chopper amplifiers eliminate 1/f noise. Figure 1-1 and Figure 1-2 compare a traditional amplifier noise spectral density to a zero-drift amplifier.
The low offset voltage of chopper amplifiers makes them an excellent choice for applications requiring high DC precision. However, as with most innovations, there are some tradeoffs that limit the effectiveness in certain applications. The goal of this document is to show the limitations of zero-drift amplifiers so that you can make an informed decision whether a zero-drift amplifier is the right choice for your specific application.
Table 1-1 compares the DC specifications like Vos and Vos drift of zero-drift amplifiers to traditional amplifiers with the best-in-class DC performance. The offset voltage of the zero-drift device is on average two to five times better than a traditional precision amplifier. The traditional devices used in this comparison are the best-in-class for DC precision and use package or laser trim to achieve the precision specification. Many other traditional amplifiers can have offsets much greater than the examples in Table 1-1 (hundreds of microvolts). The Vos drift of the zero-drift amplifiers is often tens or even hundreds of times better than traditional counterparts. The excellent stability of offset over temperature is the is the greatest advantage of zero drift amplifiers.
Table 1-1 also compares the 0.1Hz to 10Hz noise. Broadband noise is inversely correlated to the quiescent current. Therefore, when comparing noise between two different amplifier topologies, comparing devices with similar quiescent current is the best practice. With this in mind the comparison shows a significant low frequency noise advantage for zero-drift devices.
Amplifier | Feature | Tech. | Max Vos (μV) | Max Vos Drift (μV/°C) | IQ TYP (mA) | 0.1Hz to 10Hz Noise (μVPP) | TYP PSRR (μV/V) | TYP CMRR (dB) | TYP Aol (dB) | EMIRR at 100MHz (dB) |
---|---|---|---|---|---|---|---|---|---|---|
OPA392 | e-Trim™ | CMOS | 10 | 0.6 | 1.22 | 2.0 | 0.5 | 120 | 132 | 29 |
OPA277 | Laser trim | Bipolar | 20 | 0.15 | 0.79 | 0.22 | 0.3 | 140 | 140 | 38 |
OPA206 | e-Trim™ | Bipolar | 25 | 0.5 | 0.22 | 0.2 | 0.05 | 140 | 132 | 45 |
OPA928 | e-Trim™ | CMOS | 25 | 0.8 | 0.275 | 1.4 | 0.3 | 140 | 134 | 27 |
OPA191 | e-Trim™ | CMOS | 25 | 0.8 | 0.14 | 1.4 | 1.0 | 140 | 120 | 27 |
OPA210 | Laser trim | Bipolar | 35 | 0.5 | 2.20 | 0.09 | 0.05 | 168 | 132 | 35 |
OPA189 | Zero-drift | CMOS | 3 | 0.02 | 1.30 | 0.1 | 0.005 | 168 | 170 | 63 |
OPA182 | Zero-drift | CMOS | 4 | 0.012 | 0.85 | 0.119 | 0.005 | 168 | 170 | 55 |
OPA388 | Zero-drift | CMOS | 5 | 0.05 | 1.70 | 0.14 | 0.1 | 138 | 148 | 41 |
OPA333 | Zero-drift | CMOS | 10 | 0.05 | 0.017 | 0.3 | 1.0 | 130 | 130 | 65 |
OPA187 | Zero-drift | CMOS | 10 | 0.015 | 0.10 | 0.4 | 0.01 | 145 | 160 | 49 |
OPA186 | Zero-drift | CMOS | 10 | 0.04 | 0.09 | 0.075 | 0.02 | 134 | 148 | 51 |
OPA188 | Zero-drift | CMOS | 25 | 0.085 | 0.425 | 0.25 | 0.075 | 146 | 136 | 48 |