Common mode rejection ratio

Before introducing CMRR, it is necessary to reconsider the concept of input common mode voltage. As discussed in the first article of this series, the input common mode voltage of an amplifier is defined as the average voltage shared by two input terminals. This is a very convenient voltage splitting method. This can be achieved by defining Vid or differential input voltage. In current sensing applications, this can also be considered as shunt voltage. Figure 1 provides another definition of input common mode voltage that introduces differential input voltage. Figure 1 also reintroduces the concept of differential mode gain (Adm). The ideal output of a differential amplifier is the product of the differential input voltage and the differential mode gain.

Figure 1: Alternative Definition of Common Mode Voltage [2]

CMRR will affect the accuracy of current sensing solutions. It is a measure of a device's ability to suppress common mode signals. This is important because common mode signals can be displayed as differential signals inside the device, thereby reducing the accuracy of the solution. MRR is usually expressed in linear proportion in product data sheets? V/V) or logarithmic scale (dB). If reported in dB, the worst-case value is the minimum value. If expressed in µ V/V, the worst-case value is the maximum value. To calculate the error caused by the CMRR specification of the device, we need the following information: the worst-case CMRR specification in the data sheet, the common mode voltage test conditions in the data sheet specification table (Vcm pds), and the common mode voltage of the system (Vcm sys).

For example, suppose we have a system with a common mode voltage of 50V (Vcm sys) and a nominal parallel voltage of 5mV. Let's use INA170 to calculate the error, with a worst-case CMRR specification of 100dB (min) and Vcm pds=12V.

Since the standard is measured in decibels, we need to convert it to a linear scale, as shown in Equation 4. Now we calculate the error as shown in equation 5. To reduce the error contribution caused by CMRR, we have two options: increasing the parallel voltage or choosing a device with better CMRR performance. Changing the Vcm system is usually not a realistic choice as it is determined by the application. The processing method of CMRR aims to enable readers to quickly and effectively understand how it affects measurement accuracy [1,3,4].

Power suppression ratio

PSRR is a measure of the variation in Vos caused by changes in power supply voltage. The error caused by PSRR can be calculated in a manner similar to CMRR.

To calculate the error caused by the PSRR specification of the device, we need the following information: the worst-case PSRR specification in the data sheet, the power supply voltage test conditions in the data sheet specification table (Vs pds), and the power supply voltage (Vs sys) that will power the devices in the system. For example, the worst-case PSRR specification for INA170 is 10? V/V (maximum value), Vs pds=5V. If the device actually provides 30V (Vs sys), the error caused by PSRR is calculated as shown in Equation 6. As with the previous example, we assume a shunt voltage of 5mV. To reduce the error contribution caused by PSRR, we have two options: increasing the parallel voltage of the current sensor or choosing a device with better PSRR performance. Changing Vs sys is usually not a realistic choice as it is typically determined by the application. In this specific example, PSRR has been specified in µ V/V. If the value is in dB, it must be converted to linearity before applying equation 6.