Hi,
I already spent a lot of time reading other/same cases which were reported in this forum, but it looks like that you didn't have a clear answer to them yet.
I spent some days trying to fix the problem by some minor changes, such as changing the feed-forward compensation networks and etc., since our board is already in mass production now and we need to find a solution that is implementable by minor modifications. Unfortunately, this ripple never showed up in our prototypes in the past and now we are seeing it the first time in our mass production. Even in this same production, some of the boards have no problem, so the problem is non-deterministic!
Details about the problem: a TPS65130 is used in our board to produce -12V and +12V from a +5V input. The pins PSP and PSN are tied to 5V as recommended by the reference designs to activate the power-save mode. The -12V is relatively good, but the +12V has a big ripple and it's inductor is producing audible noise. The ripple is shown in the following picture:
After we patched the board and disabled the power-saving mode, the ripple disappeared. ( *No other solution such as changing the feed-forward capacitors, adding a 1M ohm resistor in parallel to the capacitor on CP, replacing the chip with an older version that we had in our inventory, replacing the inductors with different ones, changing the input and output capacitance and etc fixed the problem by the way.)
The main question: Is PSP (pin number 9 of the chip) internally pulled down or pulled up? If it is pulled down, do you think that it is possible to leave this pin floating, in order to disable the power-save mode in a reliable way? (*we have already produced many PCBs and leaving the pin floating is the only patch that we can do. Connecting the pins to GND is not possible on this PCB.)
Is there any other solution that you would recommend?
