We have a boost 3V to 5V built around ADP1614ACPZ-650-R7
The image attached shows 4 output voltage changes when removing a load (using a KEL103 electronic load in CC mode).
Apparently, the output voltage overshoot upon a falling step load drastically increases when the controller get near the peak current limit (and above).
For this test, CL res is 160k, which should set the peak current limit at rougly 1.4A. Ouput voltage starts to decrease at Iout=0.63A, which agree with the calculations.
I did the same test with different CL resistor value and the behavior is exactly the same but it occurs with an input current matching the new current limit.
Inductor is not saturating (Isat=4.4A).
I tried changing the compensation network so to increase Fc (from 4.7kHz to 8kHz) but with no noticeable improvement.
Increased load of40mA translates in more than 1V overshoot! That is... destructive.
What is going on and what can I do?
By the way, input voltage increases by ~100mV during the overshoot for all test cases shown . So you can rule the input voltage variation out of the root causes.