[SOLVED] Linear regulators still work if input voltage is too low, correct?

[cdevidal]

Sanity check: If I have a 3.3V LDO such as this one with a 300mV voltage drop and the input voltage drops to 3.4V, the output would be Vin - Vdrop = 3.1V. Correct? I don't see anything about a low-voltage lockout on this LDO.

[golden_labels]

Below the minimum supply voltage the op-amp, the voltage reference and the CE input buffers are not guaranteed to operate. So what voltage you obtain becomes unpredictable. Likely it will be below your calculated 3.04V (note: not really, see bdunham7’s comment below!), but how much below one can’t predict. It may as well turn off completely, if the CE buffers output low. Or if thermal protection pulls the gate too low.

[cdevidal]

Got it--don't count on it.

Now that I think about it, I probably don't even need to worry about this. The project is mating a 3xNiMH battery pack to an ESP32, with a lowest acceptable voltage of 3.0V, typical current of 100-200mA. Discharge curve on NiMH gives 90+% battery life at 200mAH when the cells drop to 1.2V so a regulator with a 360mV drop would do just fine.

[bdunham7]

Sanity check: If I have a 3.3V LDO such as this one with a 300mV voltage drop and the input voltage drops to 3.4V, the output would be Vin - Vdrop = 3.1V. Correct? I don't see anything about a low-voltage lockout on this LDO.

Typical behavior for LDOs is that if V_IN starts going below V_OUT plus the dropout voltage, the output starts to droop and becomes approximately V_IN minus the dropout until some point where it gives up altogether and what happens then varies.  In this case, the regulator has an ENABLE pin that requires 1 volt to turn the regulator on, so presumably it would turn off at that point if it is still going.  This series goes down to 1.0V, so it is possible it would continue to conduct reasonably well down to that point, although your load device would probably have quit long ago.  Of course you'd want to actually test its operation if any of this was critical.  Edit--looking further, it appears that the dropout voltage would start to rise when the input goes below 2.0V or so. 

Keep in mind that the 320mV V_DO is the datasheet limit for maximum current of 600mA.  Typical dropout voltage is much, much less at lower currents.  Unless you are maxing this thing out current-wise, it will probably be nearly as good as you can get without a buck-boost arrangement.