I have a battery powered system in which the battery is connected to a buck/boost converter to derive 3.3V from two or three AA batteries. The regulator I'm using is a TI TPS63050, and it has a power good output signal, which behaves as follows according to the datasheet:
The power good goes high impedance once the output is above 95% of the nominal voltage, and is driven low once the output voltage falls below typically 90% of the nominal voltage.
I'm currently using an ample bench-top power supply (with the current limit set at 1A, well above what I need) to evaluate the basic operation of the power front-end.
What I'm seeing is that, when my system (a wifi-connected ESP32-based processing subsystem) starts running I see the power good signal drop out briefly occasionally on an oscilloscope. I have a little capacitance 10uF in parallel with 0.1uF on the input to the TPS63050, and 10 uF in parallel with 10uF right at the output of the TPS63050 as well.
Is that just way too little? Looking at the input power to the scope there are some dips on the input power corresponding to dips in the output power, and correlated with the power good signal dropouts. I'm wondering if I need to put a significant bulk capacitor on the input side of the TPS3050. I don't really understand why the load I'm dealing with (a few hundred milliamps at most) would cause the bench top power supply to break a sweat instead of holding the input voltage rock-steady.
I am running power to the board through a decent length (~3 feet) of alligator-clip banana-plug cable and then a short length of Dupont-style jumper wires (~8 inches) to mate to pin headers on the board under test.