Power supply rejection ratio (PSRR) describes how well an operational amplifier (op amp) rejects variations in power supply voltage. PSRR is defined as the shift in offset per volt of shift in power supply, and generally is given units of micro-volts per volt (uV/V).
The OPA209, for example, has a typical PSRR of 0.05uV/V. Thus, for the OPA209, the power supply can change 1V with only 50nV of shift in offset (see Figure 1). This error is insignificant compared to the typical offset (35uV). Furthermore, supplies in precision systems generally have less than a volt of variation. This might lead you to conclude that you can ignore errors from power supply variation for devices with good PSRR, like the OPA209. The problem is that the specification in the data table is the dc PSRR, and often the ac PSRR is the limiting factor.
OPA209, OPA2209, OPA4209
Input Offset Voltage
Vs = ±15V, VCM = 0V
Vs Power Supply
Vs = ±2.25V to Vs = 18V
Figure 1: Input offset and PSRR Specification for OPA209
Figure 2 shows the ac PSRR curve for the OPA209. You’ll often find a high-frequency noise signal superimposed on the dc supply voltage. This high frequency noise may be the ripple of a switching power supply. The low frequency PSRR is very good, which improves the rejection of low-frequency power supply variations. On the other hand, high-frequency power supply noise rejection is poor because PSRR is low at high frequencies.
Figure 2: PSRR vs. Frequency for OPA209
The simulation results in figure 3 illustrate how the reduced PSRR at high frequencies introduces errors. In this example, a 10mV, 100kHz signal is superimposed on the power supply (Vs = 15V ± 10mV). The power supply noise is attenuated by 45dB PSRR at 100kHz and reflects as an input offset voltage variation. Notice that the output has the power supply noise and the input signal superimposed on each other.
Figure 4 gives the hand calculations for the circuit in figure 3. Note that the hand calculations compare well with the simulated results (Simulated output noise = 11.9mVpp, Hand Calculated output = 11.2mVpp).
Figure 3: Application example showing PSRR at high frequency
Figure 4: PSRR calculation for application example
In general, it is important to consider the effects of PSRR over frequency, because the PSRR specification degrades at higher frequencies. You can apply the same principle to common mode rejection. Knowing this, I hope you aren’t taken by surprise by high-frequency PSRR.
Thanks Art for article. Just small information for V(os) , G(CL) = (1+R2/R1) =101. For V(in) G(CL) = -R2/R1 , Regards.
You are correct. the offset source will see the non-inverting gain.
This is nice and i learnt something new.
Yeah, quite a good article
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.