I want to find the loop gain when two LDO outputs are connected. (@Spectre)

[dressler6]

Hi. all~
It is about stb simulation and AC simulation of Specter.

There is buffer composed of 2 stage OTA.
- DC gain 60dB
- 0dB Freq.=1Meg
- PM=60'

If these two buffers are composed as shown in Fig.1 and frequency stability is evaluated, the loop gain using STB method is 0dB. However, VOUT is almost VREF.
In Fig.2 the loop gain using RC filter method (with AC=1) becomes about 20dB. However, VOUT is also almost VREF.
(Discussion of frequency stability is excluded.)

Please explain why these results were obtained.

[Figure 1]

[Figure 2]

 

[nitishn5]

You have two feedback amplifiers driving the same point (Vout).
In this case we need to check which one of them is stronger.
Are both your buffers identical? What about the PMOS transistors?
Let us assume that the bottom amplifier and PMOS are ideal. Then it will always drive Vout=Vref no matter what the top amplifier is doing.

In the attached figure, do the analysis and see for yourself what would happen when you break one of the loops (at the arrows) while the other loop is still intact. And what is the consequence of the different gains and bandwidths. For the sake of simplicity you can just combine the stages into a single gain and a single pole.