badcaps in a ATX PSU

[RukyCon]

Re: badcaps in a ATX PSU

Click on the link from the original post, not the quote.

You also don't need 10V caps for 5V and 3.3V rails. Since you want to get d10mm caps and want to keep 3300, get these:
http://www.digikey.com/product-detai...840-ND/4843650

thank you these look like they should work

[RukyCon]

Re: badcaps in a ATX PSU

ok now all the caps in my cart are

chemi-con
2200uf 10v KY
680uf 16v KY
2200uf 16v KYB
3300uf 6.3 KYB
1000uf 16v KY

[Pentium4]

Re: badcaps in a ATX PSU

Looks good. Since they're so cheap, might as well get the little caps too At least get a startup cap. Can you read the values of the other small caps in there?

[RukyCon]

Re: badcaps in a ATX PSU

Looks good. Since they're so cheap, might as well get the little caps too At least get a startup cap. Can you read the values of the other small caps in there?

[22uf 50v] [2.2uf 50v] [10uf 50v]

[Pentium4]

Re: badcaps in a ATX PSU

2.2uF 50V: https://www.digikey.com/product-deta...1887-ND/589628

10uF 50V: https://www.digikey.com/product-deta...852-ND/3562878

22uF 50V: https://www.digikey.com/product-deta...7-1-ND/4319987

There aren't any 1uF 50V, 4.7uF 50V, or 47uF 50V in there?

[RukyCon]

Re: badcaps in a ATX PSU

that is what i can see i will try to look deeper

new ones 0.22uf 50v - 1uf 50v

i look at all the yellow caps and that's all i can find

[Pentium4]

Re: badcaps in a ATX PSU

0.22μF 50V: http://www.digikey.com/product-detai...1330-ND/589071

1μF 50V: http://www.digikey.com/product-detai...4-1-ND/4319975

[RukyCon]

Re: badcaps in a ATX PSU

my cart is now

chemi-con
2200uf 10v KY
680uf 16v KY
2200uf 16v KYB
3300uf 6.3 KYB
1000uf 16v KY

Nichicon
2.2uf 50v UPW
22uf 50v UPW
0.22uf 50v UVZ
1uf 50v UPW

rubycon
10uf 50v YXF

thanks to all that help me find new caps for my ATX power supply unit

[Pentium4]

Re: badcaps in a ATX PSU

Looks good. You're welcome. Keep us updated

[stj]

Re: badcaps in a ATX PSU

Click on the link from the original post, not the quote.

You also don't need 10V caps for 5V and 3.3V rails. Since you want to get d10mm caps and want to keep 3300, get these:
http://www.digikey.com/product-detai...840-ND/4843650

oh hell no, that's not true.

the 1.5x rule exists for a reason.
a psu output after rectification is high enough for the 5v line to blow 6.3v caps.
you can use them on the 3v3 line and on motherboards where the power is already stabilised.
but NOT on the output of a diode or rectifier!

[RukyCon]

Re: badcaps in a ATX PSU

ok i will do so

fun fact
i had a 100% dead PSU from a gateway computer that was sadly unfixable

tho it was a crummy PSU

i saved all of the caps

all of the cap brands [that i got out of the dead PSU are]
OST
Ltec
elite
and
Capxon

[RukyCon]

Re: badcaps in a ATX PSU

oh hell no, that's not true.

the 1.5x rule exists for a reason.
a psu output after rectification is high enough for the 5v line to blow 6.3v caps.
you can use them on the 3v3 line and on motherboards where the power is already stabilised.
but NOT on the output of a diode or rectifier!

i did not place the order yet so i can go back and change the order

[Pentium4]

Re: badcaps in a ATX PSU

Huh, I've never had 6.3V caps bulge on the 5V or 5VSB rails. Isn't that why the caps have a surge voltage rating?

[Wester547]

Re: badcaps in a ATX PSU

oh hell no, that's not true.

the 1.5x rule exists for a reason.
a psu output after rectification is high enough for the 5v line to blow 6.3v caps.
you can use them on the 3v3 line and on motherboards where the power is already stabilised.
but NOT on the output of a diode or rectifier!

Not sure where the "1.5x rule" stems from. It doesn't make sense. And according to Panasonic:

The life of the aluminum electrolytic capacitor is influenced by an ambient temperature, a ripple current, and an applied voltage. Especially, the ambient temperature and the ripple current are the main factors that causes wear-out failure. Applied voltage does not influence life time too much if an applied voltage to the aluminum electrolytic capacitor is below0.9x the rated voltage.but can be a factor of sudden failure.

I realize that the endurance tests have the caps in series with 10-1000Ω resistors so as to make certain the peak voltage does not exceed the applied voltage, but if the 1.5x rule really held true, there wouldn't be so many PSUs that would get away with using 6.3V rated capacitors on the +5V output or 16V rated capacitors on the +12V output (same goes +12V input capacitors on motherboards). Many have gone by a "1.2x" rather than "1.5x" ruling with no issues. Quality capacitors will be able to handle their rated voltage (and even the surge voltage for a few seconds) with no issue (as they do for thousands of hours at 105ºC, with the rated ripple current applied in the endurance tests).

[RukyCon]

Re: badcaps in a ATX PSU

so should i use a 6.3v cap or a 10V cap?

[stj]

Re: badcaps in a ATX PSU

use a 10v cap

for everybody else:
remove the caps on the 5v line of a psu and scope the output voltage - you'l be surprised.
and to re-use a quote:

Applied voltage does not influence life time too much if an applied voltage to the aluminum electrolytic capacitor is below0.9x the rated voltage.but can be a factor of sudden failure

unsmoothed 5v can in theory be 7.05v
now that looks like a lot more than 0.9 x 6.3v to me!

[RukyCon]

Re: badcaps in a ATX PSU

ok i will do that

[RukyCon]

Re: badcaps in a ATX PSU

so far all of the 3300uf 10v cap i found are 12.50mm wide i need a cap that's 10mm wide.

i found a rubycon PX cap that's 10mm wide but i don't know if the rubycon PX caps can be used in a PSU

can someone help me
thank you

[stj]

Re: badcaps in a ATX PSU

i see what you mean, closest i can see is panasonic FK 2700uf

[Wester547]

Re: badcaps in a ATX PSU

use a 10v cap

for everybody else:
remove the caps on the 5v line of a psu and scope the output voltage - you'l be surprised.
and to re-use a quote:


unsmoothed 5v can in theory be 7.05v
now that looks like a lot more than 0.9 x 6.3v to me!

Well, electrolytic capacitors work such that the thickness of the anode oxide layer is proportionate to the voltage applied to the plates. So if you apply 5V to a 5V capacitor, it is essentially a 5V capacitor and no longer a 6.3V capacitor. I'm pretty sure the magamp circuit (toroidal output inductor) in a PSU before the caps play a part in regulating and smoothing the rectified output voltage (the function of the larger inductors is similar to that of smaller coils in motherboard buck regulators, but if the capacitors go open circuit or high-ESR, that can mess up the output voltage). In discontinuous flyback topology (and continuous, which is less stressful but complicates the design of the control loop), the output inductor is part of the transformer secondary.

And PX is a general purpose Rubycon series. Probably best to stick to entry-level low-impedance parts such as those mentioned (LXZ, FC, PW, KY, KYB, HE, etc). This is a half-bridge PSU so the output frequency will be lower, and the need for very low impedance parts on the output is lesser (capacitance matters more, especially since it goes down with frequency). Panasonic FK could work too.