TPS65130 problems with noise -- ripple and more.
With our circuit, ripple is on the order of 4KHz, 20 mV p-p. 5V in, +12V and -12V out. Power saving mode enabled. Loading is light, 50 mA and 20 mA respectively.
Outputs see in our circuit:
1, 22 uF cap to GND, both +12V and -12V
2, 22 uH inductor to another 22 uF cap to GND, whence the inductor feeds the board's DAC and op amp loads
3, DAC and op amp loads on +12V, with about 50 uF total distributed decoupling (in addition to the 22 uF in item 2 above)
4, Op amp loads on -12V, with about 3 uF total distributed decoupling (in addition to the 22 uF in item 2 above)
5, No feedforward capacitors on feedback resistors
6, CN = 4.7 nF and CP = 10 nF
Experiment:
Added about 50 uF and 20 uF respectively to +12V and -12V decoupling around the board
Result: worse noise. Now sees ~100 mV bumping in the 10 Hz range, and periodically varying ripple amplitude.
Question:
1, In the absence of a Spice model, how much output capacitance can the TPS65130 handle at light loads and with power saving enabled, without ripple increasing?
2, What is the likelihood that power saving mode is interacting badly here?
