Power Supply Resevoir Size

[DF96]

You won't find many textbooks in the fiction section.

[OnAudio]

How about a book on alternative technology in the scifi section

[Kindhornman]

Harrison,
Curious mind want to know. I do see ideas from Star Trek, so we all know that sometimes an idea is just before its time....

And that reminds me I have to complete a paper on advanced composites for a seminar speech I signed on to give.

Steven

[gootee]

Originally Posted by OnAudio

Fun stuff about parallel caps

How does each parallel cap see 8 Ohms when there is only a 4-Ohm load?? Impossible. If they are paralleled across 4 Ohms then they each see 4 Ohms.

But it SHOULD be about what the 4 Ohms sees. If the caps are the same as the original, it sees about twice the capacitance, about half as much inductance (ESL), and about half as much ESR. EVERYTHING is improved.

If the two parallel caps are each half the value of the original, then the capacitance stays about the same as the original but the ESL and ESR are probably less than half as much as with one cap, because the caps are smaller and can be placed closer and their connections and lead spacings could be shorter. HF behavior is improved.

Of course, for decoupling, the caps wouldn't usually be right across the load. They'd be from a power rail to the load ground, typically.

But if it was a trick question, then "out with it"!

[OnAudio]

Originally Posted by gootee

How does each parallel cap see 8 Ohms when there is only a 4-Ohm load?? Impossible. If they are paralleled across 4 Ohms then they each see 4 Ohms.

But it SHOULD be about what the 4 Ohms sees. If the caps are the same as the original, it sees about twice the capacitance, about half as much inductance (ESL), and about half as much ESR. EVERYTHING is improved.

If the two parallel caps are each half the value of the original, then the capacitance stays about the same as the original but the ESL and ESR are probably less than half as much as with one cap, because the caps are smaller and can be placed closer and their connections and lead spacings could be shorter. HF behavior is improved.

Of course, for decoupling, the caps wouldn't usually be right across the load. They'd be from a power rail to the load ground, typically.

Hehe, I knew that would tickle you. Using I = V/R. Assuming the load is 4 ohms and the voltage is 4 volts. A single cap would release 1A. Now if there are two caps, each cap would release 0.5A . R = V/I = 8 Ohms

[OnAudio]

All the best. Enjoy the space ship building

Originally Posted by Kindhornman

Harrison,
Curious mind want to know. I do see ideas from Star Trek, so we all know that sometimes an idea is just before its time....

And that reminds me I have to complete a paper on advanced composites for a seminar speech I signed on to give.

Steven

[Kindhornman]

gootee,
Is there some theoretical point where adding more capacitors becomes a problem? Do the multiple leads between capacitors end up adding some inductive or resistive problems at some point, or could you theoretically parallel hundreds of small value caps?

Steven

OnAudio,
Just trying to teach engineers how to build better aerospace and automotive parts faster and cheaper. Spaceships would be more fun. but my dylithium crystal are on back order.

[OnAudio]

Originally Posted by Kindhornman

but my dylithium crystal are on back order.

Will send you my shipping address, needed some of that

[Tony]

Originally Posted by Kindhornman

gootee,
Is there some theoretical point where adding more capacitors becomes a problem? Do the multiple leads between capacitors end up adding some inductive or resistive problems at some point, or could you theoretically parallel hundreds of small value caps?

Walt Jung in his 1980 Audio magazine article "Build an Energy Storage Bank" mentioned this, he said at some point adding more capacitance, the lead inductance becomes significant to affect the results....i do hope he chimes in in this discussions...

[danielwritesbac]

Originally Posted by Kindhornman

Do the multiple leads between capacitors end up adding some inductive or resistive problems at some point, or could you theoretically parallel hundreds of small value caps?

Sure fire recipe for very dull audio. With massive paralleling, there seems to be some treble killing gyrations going on. That's worse than a choke. It is about as bad as my SD669 capmulti. Oh fail! But I learned that any possible audio processor effects can be installed at the power circuit because whatever you do to the power circuit also affects the output. If on purpose, there's a powerful tool, but powerfully vexing if done by accident.
P.S.
A more customary spot to find the tone control feature is at the amp board's power cap sizes--bigger for more laid back (possibly duller) -or- smaller for more forward (possibly clearer). It isn't entirely a case of no free lunch. An even balance plus extra good noise filtering can get you a combination of both laid back and clearer but at the cost of more time investment.
P.P.S.
Your massively paralleled example is not problematic if the predrive is run on regulators. Well, that's the orthodox fix. I have a cheaper method, at post 1115.