Power supply thoughts...

[PatrickPicard]

Hi,

I'm designing a new power amp and i was thinking about multiple way to design power supply... I need to know which one of these 3 I should use:

In each one, do I have to consider the formulas (τ = Rload x C) and (Xc = 1 / (2 x π x F x C))) ???

First one:

Using only 1 capacitor (25 000 uF) on each rail...

(simple, but really effective???)

Second one:

Using 4 capacitors (6 800 uF) connected in parallel on each rail...

(is it more stable or???)

Third one:

Using many capacitors (8 200uf, 6 800uf, 4 700uF, 3 300uF, 2 200uF, 1 000uF, etc...) connected in parallel on each rail...

(maybe each capacitor will filter some frequencies (Xc) by working independently and be more effecient to low and high frequencies and current peak from the power amp???)

What are your thoughts about that???

THANKS,

Pat p.

[bob91343]

It really depends on the construction of the capacitors. You are paralleling all the ESRs but that won't buy you much if the ESR is proportional to the capacitance. Look at the manufacturer's data.

You filter high frequencies better, in general, with smaller capacitors because of the reduced inductance.

Another factor is how you construct the unit. Will you put all the capacitors together in one bank, or put them here and there? What size wire will interconnect?

There are so many variables that you have to consider it as a system. And there is the 'all the eggs in one basket' idea that many parts will be statistically closer to published failure rate while just having a few parts makes statistics worthless. And if one capacitor fails open, the circuit will operate more or less normally - unless it's the only one.