... is that an LDO regulator doesn't suppress HF noise as well as the old classic linears, as they have less gain.

My 30-year old linear PSU's show about 5 mV of random (and I do mean random) noise, mostly at high frequency under significant switching load conditions, e.g. driving steppers, firing banks of LED's, etc.

Switchers PRODUCE their own noise (unwanted signal) at their outputs. Considerable additonal effort is required to manage this noise. At extremely low currents, gaining in popularity these days, it may be tolerable to use a switching supply to power a digital circuit. If, however, there are analog portions, particularly gain stages, involved, it's unlikely the process will be painless. I can't imagine getting reasonable performance from a 16-bit ADC with the power-to-gnd noise typical from switching regulators.

Kai has repeatedly shown how to go about managing this sort of problem. It's not simple and not cheap. In some cases, it's worth the extra trouble and expense.

What this article fails to address is that the input noise from mains power is of sufficiently low frequency that linear regulators, even LDO's, CAN deal with it. The better "low-noise" switchers, however, produce ripple of high enough frequency that linear post-regulators can not eliminate.

It becomes a matter of deciding which problem you wish to solve. I'd say the best you can do is devise a reasonable compromise.

RE