We are using the 3.6-V version of the LP2985-N LDO regulator in a situation where its battery-supplied input voltage will eventually fall as low as 3.4 V. As expected, the LP2985 output voltage tracks the input voltage when running open-loop. What I did not expect is a nearly order-of-magnitude increase in noise. What accounts for the noise?
From TI engineer Michael Day's informative paper, Understanding Low Drop Out (LDO) Regulators: "Entering dropout also affects an LDO’s output noise. An LDO’s output noise is reduced when the LDO is in or near dropout. All bandgap references (VREF) produce noise, and in an LDO, the bandgap noise is fed directly into the error amplifier as shown in Fig. 1. The noise on the output of the amplifier modulates the pass element’s gate voltage which generates noise on the LDO output. When an LDO is in dropout, the error amplifier output voltage is at its maximum, and its ability to pass bandgap noise from its input to its output is reduced. This reduces output noise. This feature is sometimes used with standard LDOs in a low noise application like powering a phased lock loop (PLL). If tight regulation is not needed, the LDO can be intentionally operated in dropout to provide this feature."
