power capacitors

[tuben]

Hello everybody out there

One question.

Which is best in a powersupply

Using one capacitor at 10.000 microfarads
or wiring up five capacitors at 2.000 microfarads in parallel ?


Tuben

[ble0t]

It's typically better to use the smaller ones in parallel.....performance should be better and it's most likely going to be a lot cheaper too

[tuben]

Hello.

Thanks for the info.

I had forgotten that I had about 100 2700/63 capacitors in a box in the basement, may come to use now...

But then I also have some 10.000/63 and 15.000/40 !

Decisions decisions...


Tuben

[sam9]

Supossedly, parralleling 5 2000uF's will result in a lower ESR. Whether this really results in "quieter" rails or even lower ripple is something I don't know. There are lots of opinions based on theory, but I wish someone with good instumentation would conduct some comparrissons. (Those of us with limited resources don't have the means to check this ourselves.)

The following is not what you asked but I'll offer what I can offer

A- If ordering from catalogs, multiple caps usually cost slightly more, but not always. Sometimes you can find a special source where the cost is much lower.

B- The relative dimensions (height vs. area) of the enclosure sometimes control the decision.

C- A double layer of axial caps is a possible volume-efficient configuration

D- The above are not completely unrelated to sonics as configuration permitting more total capacitance may be a possability, even if it is a brute-force approach lacking finese.

[DSP_Geek]

Originally posted by tuben
Hello.

Thanks for the info.

I had forgotten that I had about 100 2700/63 capacitors in a box in the basement, may come to use now...

But then I also have some 10.000/63 and 15.000/40 !

Decisions decisions...


Tuben

Make sure you reform the dielectric, as electrolytics which haven't seen voltage for a while have a disconcerting tendency to go POOF.

How does one reform an electrolyic? First, get a power supply sourcing the working voltage, then place a high-value resistor between this supply and the cap to be reformed. Make sure the time constant is quite long, on the order of several minutes, so for 2700 uF you're looking at 100 kohms minimum for a 270 second (> 4 minute) time constant. Let the cap charge for half an hour or so, then remove it and discharge it through 100 ohms.
Lather, rinse, repeat.

I heard the above from folks who've succeeded but I haven't done this myself, so if someone else wants to chime in, please do so.


Francois.

[Jay]

Like Sam9, IMO more caps of the same total value should be
more expensive than a single cap.

I prefer 4 caps than 1 cap, but too much? I don't like it.
IMO, one big cap have bass impact that is not comparable with
too many small caps of the same value. Well, may be it is
just power supply mud that make up the good bass but I doubt
you will prefer the many caps if you can hear the difference.

DSP_Geek, I'm not sure what chime-in means but I will chime in
anyway I think reforming the electrolityc is just some kind
of burning-in? I don't know what can be expected from doing
it. The caps will undergo the same process anyway during
operation, especially if much current is drown by the amplifier.
BUT, for very very old caps I tried lower voltage first before
the real high one. But you should be able to measure it with
ohm-meter before using it, and bad (old) caps just do not have
to be used IMO.

Oh BTW, I discharge caps by connecting with low value resistor.
Surely there will be spikes or fires. I'm not sure whether this
may damage the cap, I just need to do it fast