SBVA093 December 2022 LP2992 , TPS786 , TPS7A30 , TPS7A3001-EP , TPS7A33 , TPS7A39 , TPS7A4501-SP , TPS7A47 , TPS7A47-Q1 , TPS7A4701-EP , TPS7A49 , TPS7A52 , TPS7A52-Q1 , TPS7A53 , TPS7A53-Q1 , TPS7A53A-Q1 , TPS7A53B , TPS7A54 , TPS7A54-Q1 , TPS7A57 , TPS7A7100 , TPS7A7200 , TPS7A7300 , TPS7A80 , TPS7A8300 , TPS7A83A , TPS7A84 , TPS7A84A , TPS7A85 , TPS7A85A , TPS7A87 , TPS7A89 , TPS7A90 , TPS7A91 , TPS7A92 , TPS7A94 , TPS7A96 , TPS7B7702-Q1 , TPS7H1111-SEP , TPS7H1111-SP , TPS7H1210-SEP
Upon review of Equation 4 and Equation 5, there are no theoretical limitations on the number of LDO's that can be paralleled. Also, there is no assumption that all of the LDO's or output voltages must be the same. To simplify implementation, it may be advantageous to reduce variability by selecting the same LDO with the same nominal output voltage. The worst-case current sharing of each LDO can be obtained and is a function of the error voltage and ballast resistance chosen. As discussed in Section 2, the sources of variability for single LDO's are located on the VREF node and VFB node.
When the designer has access to the VREF node, they can short all VREF nodes together, providing an identical reference voltage for each device. The reference voltage is typically the dominant source of variability in steady state analysis, so connecting the reference voltage nodes together can make a significant improvement in the current sharing of each parallel LDO. With the references shorted we can simplify Equation 7 and Equation 8 for our current sharing analysis.
The feedback for each LDO must be before the ballast resistors and not directly on the load voltage VLOAD. Thus, we cannot tie the VFB nodes together to eliminate its contribution to VE. When selecting parallel LDO's without unity gain feedback, it is desired to minimize ∆IBIAS (listed in the LDO data sheet) or its effects on accuracy by choosing the proper feedback resistance.