Low noise power supply - linear?

[Cirkit]

Is it possible to design a low noise, low ripple power supply using a switching design with an LDO?

One typical application is for use with sensitive pre-amp circuits typically measuring pA/nA currents?

The preference by many is to just use a linear power supply.

 

[HellasTechn]

My opinion on this would be to use a linear PSU. The classic transformer with full wave rectifier and smoothing capacitor. Why would you use SMPS for such low currents? Unless if you wish to make it tiny in size.

 

[Cirkit]

I'm just thinking power consumption if it were left on continuously.

At what threshold would you consider using a switching design with an LDO?

 

[Ylli]

Definitely a linear supply. Or if you really want a quiet supply, then use batteries.

 

[davenn]

I'm just thinking power consumption if it were left on continuously.

At what threshold would you consider using a switching design with an LDO?

Maybe it would be helpful if you told us what you are doing, then you would get better answers, eg.

voltage needed
current needed
run time needed

 

[(*steve*)]

I've looked at a number of designs for pA meters. There's more issues than just ripple, you want to avoid noise coming from all sources. Batteries can be contained within a shielded enclosure, once you go to mains operation, you note need to filter the noise coming in through your power supply and deal with capacitive (or direct) coupling via the power supply.

Most pA meters don't require lots of power, and an alternative option is to use several small lipo cells in series. The charger for these will likely be noisy (as well as introducing coupling issues) so don't use it when you're charging.

There's a very simple pA meter design that runs from a 9v battery and has a resolution down to 100's of fA. Have you seen that?