Type of capacitors to use for psu?

[bord]

I've read the FAQ/threads but came across this reviewer who recapped 2 psu's using Fuhjyyu and Capxons with chemicons KZE/KY but ended up with worse results.

Could it be that some psu's don't do well with low esr caps depending upon the design?

[selldoor]

Re: Type of capacitors to use for psu?

You need to put links in to the relative threads/posts so we can see what you refer to. Cant even tell if you mean for instance an ATX PSU or a monitor power supply. Maybe then the member/s who wrote the original will be able to comment.

[bord]

Re: Type of capacitors to use for psu?

You need to put links in to the relative threads/posts so we can see what you refer to. Cant even tell if you mean for instance an ATX PSU or a monitor power supply. Maybe then the member/s who wrote the original will be able to comment.

I wasn't purposely being cryptic but the web link was stripped out of my post and I can't edit my post possibly as a measure to reduce spam.

So I'll have to spell out the title of the article so you guys can google it - "Capacitors and the Computer PSU" by jonnyguru.
https://www.jonnyguru.com/modules.ph...tory&ndar_id=8

As you can see on page 4, the measured ripple of the recapped psu's became worse after replacing them with Chemicons KZE/KY. I do remember some decent psus using mediocre cap brands like OST which would not normally survive well on motherboards but seem do do ok on psus. Is it because psu's work better with non low-esr caps?

[mariushm]

Re: Type of capacitors to use for psu?

That particular power supply was "optimized" for capacitors with slightly higher esr than the esr of those KZE capacitors. By lowering the ESR, the feedback to the main controller changed, so now the power supply runs at a slightly frequency which can affect the ripple.
Notice that the old Fu*$^$ capacitors are also very tall, and the height can affect other things, perhaps higher inductance.
Also, you may notice the ton of solder at the bottom.. even that can affect measurements.

Yeah.. generally you want to replace capacitors with something that has equal or slightly lower esr, but you don't want to get the lowest esr possible if the power supply used crappier capacitors.

Even if you could, you might now want to use ultra low esr capacitors because those are usually water based and could dry out/ degrade fast in places like a power supply (squeezed between hot inductors and covered by cables so hard for air from fan to cool them)

In newer gold/platinum efficiency though, most designs like as low esr as possible.

[bord]

Re: Type of capacitors to use for psu?

That particular power supply was "optimized" for capacitors with slightly higher esr than the esr of those KZE capacitors. By lowering the ESR, the feedback to the main controller changed, so now the power supply runs at a slightly frequency which can affect the ripple.
Notice that the old Fu*$^$ capacitors are also very tall, and the height can affect other things, perhaps higher inductance.
Also, you may notice the ton of solder at the bottom.. even that can affect measurements.

Yeah.. generally you want to replace capacitors with something that has equal or slightly lower esr, but you don't want to get the lowest esr possible if the power supply used crappier capacitors.

Even if you could, you might now want to use ultra low esr capacitors because those are usually water based and could dry out/ degrade fast in places like a power supply (squeezed between hot inductors and covered by cables so hard for air from fan to cool them)

In newer gold/platinum efficiency though, most designs like as low esr as possible.

You're right about newer psus since those 2 were older psus and the Antec wasn't even 80+.

I just read the there was discussion thread I the article I posted above. I missed that the division marks on the graphs were 0.01v so a few hundredths off isn't going to have a practical difference. And one poster said that the new caps needed time to break in which could explain a few jittery up and down results. So I don't think it made things worse even if it wasn't measurably better.

What I didn't get was the comment that "Neither the KZE or KY series are particularly low impedance". What does that mean? I though low esr/low impedance followed each other.